I&EPA
United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park NC 2771 1
EPA-600- 2-80-026
January 1980
| C ID Research and Development
ENV1RON!W^TAF
AG " Y
DALLAS, I£;
Portable Miniature
Sampler for Potential
Airborne Carcinogens
in Microenvironments
Phase 1
Development
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-80-026
January 1980
PORTABLE MINIATURE SAMPLER FOR
POTENTIAL AIRBORNE CARCINOGENS IN
MICROENVIRONMENTS
Phase 1. Development
J. J. Brooks and D. S. West
Monsanto Research Corporation
Dayton, Ohio 45407
Contract No. 68-02-2774
Project Officer
James Mulik
Atmospheric Chemistry and Physics Division
Environmental Sciences Research Laboratory
Research Triangle Park, North Carolina 27711
ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
-------
DISCLAIMER
This report has been reviewed by the Environmental Sciences
Research Laboratory, U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute endorsement or recommendation
for use.
11
-------
ABSTRACT
A 3-year research program was initiated for the purpose of
developing a portable, miniature, sorbent-type collection system
for sampling and preconcentrating organics in general, and car-
cinogens and associated compounds (e.g., mutagens, precarcinogens,
and cofactors) in particular, from ambient air. The purpose of
such a system is to assess the exposure of individuals and/or
small groups of individuals to these types of compounds in vari-
ous environments. Inherent in the ability to assess exposures is
not only the sampling capability but also analytical confirma-
tion. The determinative step in this program will be capillary
gas chromatography/mass spectroscopy.
Progress during the first year is discussed and deals with the
selection of candidate sorbent materials; the selection of test
compounds for sorbent evaluation; the evaluation of the sorbent
materials in terms of capacity, desorption properties, and physi-
cal properties that relate to pressure drops and ultimate system
design; the selection of a three-sorbent system based on Tenax-
GC, Porapak R, and Ambersorb XE-340; and the consideration of
various design parameters for the pump and system hardware.
The selection of Tenax-GC, Porapak R, and Ambersorb XE-340 was
based on the need to have a system that would be capable of sam-
pling a broad range of organic compounds in terms of volatility
and polarity and still be amenable to recovery processes, partic-
ularly thermal desorption. Other sorbents evaluated during this
process were Porapak N and SKC activated coconut charcoal. The
111
-------
test compounds used to evaluate sorbent characteristics included
n-butane, benzene, propylene oxide, benzyl chloride, acrylonitrile,
phenol, iso-octane, naphthalene, bis(2-chloroethyl) ether, 1,2,4-
trichlorobenzene, ethylene glycol, m-nitroanisole, n-hexadecane,
phenanthrene, hexachloro-1,3-butadiene, 4-bromodiphenyl ether,
succinotrile, and o-nitroaniline.
This report was submitted in partial fulfillment of Contract No.
68-02-2774 by Monsanto Research Corporation under the sponsorship
of the U.S. Environmental Protection Agency. This report covers
a period from 30 September 1977 to 30 September 1978, and work
was completed as of 30 September 1978.
IV
-------
CONTENTS
Abstract iii
Figures vi
Tables viii
1. Introduction 1
2. Conclusions and Recommendations 2
3. Technical Progress - Phase I 3
Task 1 - Initial research 3
Task 2 - Sorbent evaluation 14
Task 3 - Design of the portable miniature
collection system prototype 57
Additional work - analytical development 65
References 75
Appendices
A. Standard sample generation system 76
v
-------
FIGURES
Number Page
1 Capacity data for selected compounds on Tenax-GC
superimposed on boiling point vs. dipole moment plot . 6
2 Capacity data for selected compounds on Tenax-GC super-
imposed on boiling point vs. relative polarity plot. . 10
3 Capacity data for selected compounds on Chromosorb 104
superimposed on boiling point vs. relative polarity
plot ......................... 11
4 Elution analysis method for sorbent capacity
determination ..................... 16
50^
5 Plot of log °Vg vs. 1/T for selected test compounds
on Porapak R ..................... 18
Plot of log Vg vs. 1/T for selected test compounds
on Porapak R ..................... 19
Plot of log Vg vs. 1/T for test compounds on
Porapak N ....................... 21
8 Plot of log Vg vs. 1/T for all test compounds on
Tenax-GC ....................... 22
9 Plot of log Vg vs. 1/T for standard gases on
Amber sorb XE-340 ................... 23
10 Plot of log Vg vs. 1/T for standard gases on SKC
charcoal ....................... 24
11 Correlation of
-------
Number
14 Diagram of six-port, two-position valve within the
Chromalytics oven ................... 34
15 Differential scanning calorimetry data for Chromosorb
102 .......................... 42
16 Thermogravimetric analysis data for Chromosorb 102 ... 42
17 Desorption/analytical system used in tube background
evaluations ...................... 46
18 Experimental chromatograms of the backgrounds of a
Tenax-GC sampling tube before and after conditioning . 47
19 Pressure drops vs. flow rates for 15 cm long, 0.6 cm
O.D., 4 mm I.D. glass sampling tubes ......... 52
20 Specifications of selected sampling tube design ..... 50
21 Plots of requirements for various sampling arrangements
and the manufacturer's specifications for four
miniature pumps .................... 60
22 Plots of requirements for various sampling arrangements
and the manufacturer's specifications for two large
(desk top) pumps ................... 61
23 Sketch and description of small personnel sampler. ... 63
24 Sketch and description of large desk-top sampler .... 64
25 Two chromatograms of JP-4 jet fuel ........... 66
26 Flow schematics of two GC modifications ......... 68
27 Capillary GC with capillary inlet system ........ 69
28 Flow schematic of capillary inlet system ........ 70
29 Injections of JP-4 into a 0.8 mm (0.03 in.) diameter,
liquid N2 cooled capillary trap, and desorption of
trapped sample into a SF96 WCOT column for analysis. . 72
30 Evaluation of capillary inlet system with samples of a
six-component mixture (pentane, hexane, benzene, hep-
tane, toluene, and octane), collection in liquid N2
cooled trap, and desorption into a SF96 WCOT column
for analysis ..................... 73
vn
-------
TABLES
Number Page
1 Compounds Included in Volatility Polarity Plots 7
2 "Scoring" System for Polarity Evaluation 8
3 Test Compound Matrix 13
4 Matrix of Aliphatic Test Compounds 15
5 Matrix of Aromatic Test Compounds 15
6 Retention Times and Capacities for Selected Test Com-
pounds on Porapak R 17
7 Capacities for Selected Test Compounds on Porapak R. . . 16
8 Compilation of Capacity and Solubility Parameter Data. . 25
9 Correlation Coefficients for the "Best" Lines of 6
FET m
versus Log Vg,,n0r, Plots 30
10 Sampling Tube Data (All Sorbent-Filled Tubes Have Glass
Wool Plubs) 32
11 Chromatographic Conditions and Raw Data for the Deter-
mination of Desorption Efficiencies 36
12 Retention Time Data and Desorption Efficiencies at or
Near Desorption Temperature for Sorbent Materials. . . 37
13 Qualitative Evaluation of Two Sorbent Properties Compared
to 6 to Determine Ranges of Sorbent Utility 39
14 Ranges of Utility of Solid Sorbents Based on 6 40
15 Expected Thermal Decomposition Behavior of Selected
Sorbent Materials 44
16 Pressure Drop Data for Glass Sampling Tubes 51
17 Adsorbent Properties Chart 54
Vlll
-------
Number Page
18 Evaluation of Sorbent Package for Two Modes 56
19 Results of Pressure Drop Studies Across Two Diameters
of Tapered Glass Sampling Tubes 58
20 Total Pressure Drops Across Different Sampling
Configurations 59
IX
-------
SECTION 1
INTRODUCTION
A wide variety of gases, vapors and particulates of organic com-
pounds, ranging from volatile hydrocarbons (e.g. methane) to high
molecular weight phthalates (plasticizers), polychlorinated
biphenyls (pesticides), and polynuclear aromatics, are present
within the atmosphere. Some of these organic compounds have
been shown or are suspected to be carcinogens, mutagens, cofac-
tors, or precursors and therefore are of environmental and health
concern. The objective of this three-year project is to develop
a portable miniature collection system capable of concentrating
a wide variety of organic compound vapors from ambient air. This
system is then to be evaluated for assessing the exposure of an
individual or small group of individuals to carcinogens, mutagens,
cofactors, precursors, and the like.
The text of this report focuses on the work conducted during
the first year of this project. This work included the survey,
evaluation, and selection of adsorbent materials for use within
the portable miniature collection system and the initial develop-
ment of this collection svstem.
-------
SECTION 2
CONCLUSIONS AND RECOMMENDATIONS
Five solid sorbent materials were selected from the variety of
sorbent materials commercially available based upon information
obtained through a literature survey. These sorbents were then
evaluated with a matrix of eighteen test organic compounds selec-
ted to represent a wide range of volatilities, polarities, and
functionalities. Based upon laboratory evaluations of the capa-
city, desorption efficiency, background, decomposition, and pres-
sure drop properties of these five sorbents, a combination of
three was selected for use within a portable miniature air
collection system. The three sorbents selected were Tenax-GC,
Porapak R, and Ambersorb XE-340 which should be useful for the
collection/analysis of low volatility, intermediate volatility,
and high volatility compounds, respectively.
Tapered glass tubes of specific dimensions we,re determined to
be the best means of containing the sorbents and a basic sampling
system was designed, which consisted of three sorbent-filled glass
tubes in a series or parallel arrangement, followed by optional
flow control/measurement equipment (needle valves and rotameters),
and ending with a portable, battery-operated pump.
Future research will include an evaluation of the design options
of this basic sampling system and validation of its use in assess-
ing the exposure of an individual or small group of individuals
to carcinogens and related compounds through laborabory and field
testing.
-------
SECTION 3
TECHNICAL PROGRESS - PHASE I
TASK 1 - INITIAL RESEARCH
Literature Survey
Before initiating studies to evaluate and select sorbent mate-
rials, a comprehensive review of available literature on the use
of solid sorbent materials for sampling organic vapors was con-
ducted. Information was collected on specific sampling applica-
tions for over 110 compounds using more than 30 different sorbent
materials. The pertinent knowledge obtained from this review
included:
Capacity/efficiency information for various compounds on
various sorbents.
A variety of desorption techniques, thermal and solvent.
General knowledge of sorbent properties, capabilities, and
problems.
Sorbent conditioning procedures.
Possible experimental techniques for sorbent evaluation
and sampling validation.
-------
Selection of Candidate Sorbent Materials
At least three and not more than six commercially available
sorbent materials were to be selected for further laboratory
evaluation based on information obtained through the literature
survey. This survey indicated three major classes of sorbent
materials, which are listed below with representatives of each
class.
(1) Porous polymers - Tenax-GC
Porapaks
Chromosorbs
XAD' s
(2) Carbonaceous materials - Activated carbons and charcoals
Graphitized carbon black
Ambersorbs
(3) Others - Molecular sieves
Silica gel
Liquid coated solid supports
Generally, the porous polymers were found to have the most desir-
able properties for air sampling, including low background and
low reactivity, as well as high capacities for many compounds.
Unfortunately, the porous polymers were found to have little
capacity for the more volatile compounds. On the other hand, the
carbonaceous materials were noted to have much better capacities
for volatile compounds, but are plagued with reactivity problems
and problems associated with hydrophilicity. The other sorbent
materials generally had intermediate capacities and various
associated sampling problems, and they have not been widely
utilized in air sampling applications. Specific sorbents were
selected from these three groups of sorbent materials.
-------
The goal of this selection was to choose a collection of commer-
cially available materials that, when used in an appropriate
combination, showed promise for quantitatively trapping compounds
varying widely in volatilities and polarities. To do this it was
necessary to compare the trapping capabilities (capacities) of a
sorbent for various compounds with parameters that are indicative
of volatility and polarity. Plots were made of several volatil-
ity versus polarity parameters for a collection of compounds.
Sorbent capacities for specific compounds were then superimposed
on these plots to determine whether or not there were consistent
trends in the trapping abilities of a sorbent related to these
compound parameters.
Consistent trends in sorbent capacities were seen when related to
compound volatilities as indicated by either vapor pressure or
boiling point, with boiling point being the more convenient pa-
rameter. Unfortunately, finding a single polarity parameter that
resulted in consistent sorbent capacity trends was not as easy.
For example, Figure 1 is a plot of boiling point versus dipole
moment for the compounds listed in Table 1. When capacity data
(numbers next to the circled points in Figure 1) were superimposed
on this plot, there were discontinuities in the data so that no
clear trend was evident for sorbent capacities. It was felt that
the inconsistencies in relating capacities to compound parameters
in this plot lay in an inadequate method of expressing polarity.
Therefore, other parameters (solubility parameter, hydrogen-
bonding index, and polarity index) that have been used as measures
of polarity were evaluated. Plots similar to Figure 1 were made
in an attempt to relate volatility and capacity data to each of
these parameters, but no good correlation was found for any
single parameter.
It was possible, however, to make a qualitative evaluation of the
relative merit of each of these parameters for relating the
-------
400
300. -T
200
0°
£ ioo
o
o
CO
0--
-100
LOW
POLARITY
C3
B3
Tl2
66
,
(^_ .
63~"~"
G2
E2
61
Tl
57 & 01
56 'MI
U
* 11
HI Gl
Fl .
El
43
.
C1
~"^~\ *
_A1 60
24
23
-17- ' "
(*£l2)
10 ' f
-8 \
7
-
1
MEDIUM POLARITY
W2
67
^^^_ - ' '
02 __ '
L_ '
M2
'F2 . .
B2' -A2 c* 62 D2
XI C
HIGH POLARITY mnmim
(LOW VOLATILITY)
A3
Z2 *Y2
V*2 X2 ^_______-- ""
T2*- f .
I R2 S2
65
P2
N2
K2 /2
rl^oo :*2 . MEDIUM BOILING
^ . ' (MEDIUM VOLATILITY)
1S * £«
vi' (Tm)9>5 I 59 °u wi
Ql-'Sl' R* 5'8
"4- Wi
.11. Kl
. '5° . 51 »
49 48
f^\nf\
\£>* *41
38* "3^_.
37?»33) 0 02 * 35
32* ."jl ^/ " ^-
Bl '(^29^ /^N
^o.o J^*1
^^ -^c*25
"""^ 22" (*LW ^ *
v^
15 '16 ^
V-
p8
^ -5
*4
3
1
^ "^
551; c^J 150
\,_ x
'47
* *fii
J2 -*0 _
27 ^_____ ^ "
« *36 ~~
^ ^-^ ^ '
30 - -^285 i
p ^6?O
^-.
*\/N 4
^ *18
13
LOW BOILING
(HIGH VOLATILITY)
i i i i
1.0 2.0 3.0 4.0
DI POLE MOMENT, debyes
5.0
6.0
7.0
Figure 1. Capacity data for selected compounds on Tenax-GC
superimposed on boiling point vs. dipole moment
plot (entries correspond to compounds in Table 1)
-------
TABLE 1. COMPOUNDS INCLUDED IN VOLATILITY POLARITY
PLOTS (FIGURES 1, 2, AND 3)
Key"
Compound
Key
Compound
Key
Compound
Aliphatic (and inorganic):
1 Methane
2 Nitrogen oxide
3 Methanol
4 Fluoromethane
5 Formaldehyde
6 Vinyl chloride
7 Isobutane
8 n-Butane
9 Dimethylamine
10 Neopentane
11 Cyanogen chloride
12 Nitrogen dioxide
13 Hydrogen cyanide
14 1,2-Propylene oxide
15 Methylene chloride
16 Methyl iodide
17 Carbon disulfide
18 Propionitrile
19 1,3-Propylene oxide
20 Acrolein
21 Acetone
22 Methyl acetate
23 Chloroform
Aromatic:
Al Benzene
Bl Fluorobenzene
Cl Toluene
Dl Chlorobenzene
El Ethylbenzene
Fl p-Xylene
Gl o-Xylene
HI Styrene
II Isopropylbenzene
Jl Anisole
Kl Bromobenzene
Ll Mesitylene
Ml t-Butylbenzene
Nl n-Dichlorobenzene
01 p-Dichlorobenzene
PI c-Chlorophenol
Ql Benzyl chloride
Rl o-Dichlorobenzene
24 n-Hexane
25 2,2-Dichloropropane
26 Acrylonitrile
27 Methyl ethyl ketone
28 Acetonitrile
29 t-Butanol
30 Triethylamine
31 2,3-Dichloropropene
32 1,2-Dichloropropane
33 Water
34 Nitromethane
35 3-Pentanone
36 Crotonaldehyde
37 1,3-Dichloropropene
38 n-Butanol
39 Acetic acid
40 Crotonitrile
41 1,3-Dichloropropane
42 1-Chloro-l-nitroethane
43 n-Octane
44 Valeronitrile
45 Nitropropane
46 Acetylacetone
SI Phenol
Tl p-Diethylbenzene
Ul Aniline
Vl lodobenzene
Wl Benzonitrile
XI p-Cresol
Yl Acetophenone
Zl Nitrobenzene
A2 2,6-Dimethylphenol
B2 1,2,4-Trichlorobenzene
C2 m-Chlorophenol
D2 p-Chlorophenol
E2 Naphthalene
F2 m-Dibromobenzene
G2 p-Dibromobenzene
H2 o-Dibromobenzene
12 o-Nitrotoluene
J2 m-Nitrotoluene
47 1-Chloro-l-nitropropane
48 Isoamylacetate
49 n-Butyl ether
50 l,2,-Dibromo-2-methylpropane
51 2-Heptanone
52 1-Nitrobutane
53 Cyclohexanone
54 8-Propiolactone
55 Dimethylacetamide
56 t-Butylcyclohexane
57 n-Decane
58 Ethyl acetoacetate
59 Formamide
60 Trimethylphosphate
61 n-Dodecane
62 Diethyl succinate
63 n-Tridecane
64 Dimethyl sulfone
65 Succinonitrile
66 n-Hexadecane
67 1,2,3,4,4,6-Hexachlorocyclo-
hexane
K2 m-Chloronitrobenzene
L2 p-Nitrotoluene
M2 Pyrocatechol
N2 o-Chloronitrobenzene
02 1,2,3,5-Tetrachlorobenzene
P2 m-Nitroanisole
Q2 Diphenyl ether
R2 o-Nitroanisole
S2 p-Nitroanisole
T2 o-Nitroaniline
U2 p-Dinitrobenzene
V2 .*?7-Dlnitrobenzene
W2 4-Bromodiphenyl ether
X2 m-Nitroaniline
Y2 o-Dinitrobenzene
Z2 4-Nitrodiphenyl ether
A3 p-Nitroaniline
B3 Phenanthrene
C3 m-Diphenylbenzene
The numbers in Figures 1, 2, and 3 correspond to these numbers.
order of increasing- boiling points.
Compounds are listed in
-------
capacity and volatility data to polarity. We judged the hydrogen-
bonding index to be twice as good as the dipole moment and solu-
bility parameters, while the polarity index was only about half
as good. In addition, it was noted that compounds with similar
functionalities generally had similar values for each of the
measures of polarity examined.
These observations formed the basis for a rating scheme which
gave a "composite" or relative polarity for the various function-
alities. The rating was accomplished as follows. First the
various functionalities were rated for each measure of polarity
as having a high, moderately high, moderate, moderately low, or
low value. This was done by determining the total range covered
by each parameter and qualitatively assigning a relative value
for a particular functionality depending upon where the values
for compounds having that functionality fell within that range.
Second, the measures of polarity were weighted according to how
well they correlated with the capacity data. Table 2 contains
the point value given for each combination in this weighting
scheme. Third, the point values for all the polarity parameters
were summed for each functionality to give a total "score."
These scores formed the basis for placing the functionalities on
the relative polarity scale that is used as the abcissa in
Figures 2 and 3.
TABLE 2. "SCORING" SYSTEM FOR POLARITY EVALUATION
Value
High
Moderately high
Moderate
Moderately low
Low
Hydrogen
bonding
20
16
12
8
4
Dipole
moment
10
8
6
4
2
Solubility
parameter
10
8
6
4
2
Polarity
index
5
4
3
2
1
-------
Figure 2 is a plot of boiling point versus functionality on the
relative polarity scale described above for selected compounds.
The capacities for certain of these compounds on Tenax were
superimposed in bold numbers next to the corresponding points.
Figure 3 is the same plot with capacities for certain compounds
on Chromosorb 104 superimposed. The results showed that there is
a good correlation between capacity data and a volatility/polarity
plot of this nature. This was shown by consistent trends in the
capacity data in moving vertically (i.e., from high to low volatil-
ity) or horizontally (i.e., from low to high polarity) on the plot.
It was possible to estimate lines of constant capacity. Two such
estimates are represented by the shaded areas in Figures 2 and 3.
The top shaded area in each plot represents the lower limit for
the quantitative sampling of low levels (probably ppb or less) of
compounds at a rate of 4 2,/min for 8 hours. This means that any
compound falling in this shaded area or above would probably be
quantitatively sampled under these conditions. The bottom shaded
area represents the lower limit for quantitatively sampling low
levels of compounds at a rate of 15 m£/min for 8 hours. This
shows that a slower sampling rate results in the quantitative
sampling of a broader range of compounds. It should be noted
that these were only rough approximations based on limited data
available in the literature.
Six sorbent materials were selected for further evaluation on the
basis of these and similar plots and other considerations where
there were insufficient data to construct a plot. These included
Tenax-GC, Chromosorb 104, Porapak R, Porapak N, Ambersorb XE-340,
and SKC coconut activated charcoal. Tenax-GC was selected for
the collection of moderately high to high boiling compounds,
since it has a high temperature limit. Tenax also has a slight
preference for nonhydrogen bonding compounds. On the other hand,
Chromosorb 104 was thought to prefer hydrogen-bonding compounds
which are moderately high boiling. For moderately boiling
-------
400
300
o
o
O
Q-
o
o
CO
200
100
-100
LOW POLARITY
_
_---
>.,
11
11
(.1
Fl
El*
43
Cl
40-80
A)
^^--~~
10
\
MEDIUM POLARITY
^ ^^
* B2 ~ tf 00
» M
vi
o:
M
54
kl
-300
Dl
20- 40
<0 1
\ \
HIGH POLARITY
HIGH BOILING
(LOW VOLATILITY)
V2
MEDIUM BOILING
(MEDIUM VOLATILITY)
» M2
S2
I2 D2.
. 21
-975
IP
.51
PI
14-40 ()_02_
20
9
LOW BOILING
(HIGH VOLATILITY)
i i i
HYDROCARBONS
HALOGENATED ALDEHYDES, NITRO NITRILES&t ALCOHOL & WATER
COMPOUNDS ETHERS, COMPOUNDS STRONG PHOSPHATES
ESTERS, ACIDS I
&KETONES AMINES
POLARITY
Figure 2. Capacity data for selected compounds on Tenax-GC
superimposed on boiling point vs. relative polarity
plot. (Entries correspond to compounds in Table 1.)
10
-------
400
300 -
200 -
£
o
CO
100
-100
LOW POLARITY
1
HYDROCARBONS
MEDIUM POLARITY
VI 360
[01228
234
,1169
HIGH POLARITY
HIGH BOILING
(LOW VOLATILITY)
6,090
7.1 870
840.
* PI
64 5
MEDIUM BOILING
(MEDIUM VOLATILITY)
LOW BOILING
(HIGH VOLATILITY)
i
i
i
HALOGENATED ALDEHYDES, NITRO NITRILES & t ALCOHOLS, WATER
COMPOUNDS ETHERS, COMPOUNDS STRONG PHOSPHATES
ESTERS, ACIDS I
&KETONES AMINES
POLARITY *-
Figure 3. Capacity data for selected compounds on Chromosorb 104
superimposed on boiling point vs. relative polarity
plot. (Entries correspond to compounds in Table 1.)
11
-------
compounds, the literature showed Porapak R to have the greatest
capacity, although retention time data indicated that Porapak N
(for which there were no capacity data) might actually have
greater capacity. For low boiling compounds, a wide variety of
activated carbons had been used in the literature. However, the
most common ones were those that had only moderate capacity.
Moderate capacity charcoals seemed to be necessary in order to
facilitate satisfactory desorption. Problems with activated
carbons included difficulty of cleanup, significant humidity
effects, and induced reactions of highly reactive compounds.
We decided to look at one activated charcoal, SKC coconut, and a
new product line, the Rohm and Haas Ambersorb materials. The
Ambersorbs are intermediate between the activated carbons and
polymeric adsorbents and show promising capacities for highly
volatile compounds (e.g., 81 £/g for butane) [1]. These six sor-
represent four porous polymers and two carbonaceous materials with
a variety of individual properties and the potential, in some
combination, to effectively sample compounds of a wide range of
polarities and volatilities. Later, Chromosorb 104 was eliminated
from consideration since it has been reported to have hydrophili-
city problems [2], and higher background properties than Tenax-GC
[3]. In addition, it would not be expected to add any capabilities
to those offered by the Tenax and Porapak materials.
[1] Holzer, G., H. Shanfield, A. Zlatkis, W. Bertsch, P. Juarez,
H. Mayfield, and H. M. Lieblich. Collection and Analysis of
Trace Organic Emissions from Natural Sources. Journal of
Chromatograpny, 142:755-764, 1977.
[2] Pellizzari, E. D., J. E. Bunch, R. E. Berkley, and J. McRae.
Collection and Analysis of Trace Organic Vapor Pollutants in
Ambient Atmospheres. The Performance of a Tenax GC Cartridge
Sampler for Hazardous Vapors. Analytical Letters, 9(1) :45-
63, 1976.
[3] Pellizzari, E. D., R. H. Carpenter, J. E. Bunch, and
E. Sawicki. Collection and Analysis of Trace Organic Vapor
Pollutants in Ambient Atmospheres. Thermal Desorption of
Organic Vapors from Sorbent Media. Environmental Science &
Technology, 9(6):556-560, 1975.
12
-------
Selection of Test Compounds
To evaluate the selected sorbent materials for their sampling
capabilities, it was necessary to establish a matrix of test
compounds that sufficiently described the ranges of volatilities
and polarities that were anticipated in this program. The matrix
shown in Table 3 was established as a guide to the selection of
appropriate test compounds.
TABLE 3. TEST COMPOUND MATRIX
Low
Aliphatic :
Low volatility
Medium volatility
High volatility
Aromatic:
Low volatility
Medium volatility
High volatility
polarity
X
X
X
X
X
X
Medium polarity
X
X
X
X
X
X
High polarity
X
X
X
X
X
X
The volatility/polarity plots of various compounds shown in
Figures 2 and 3 were used to choose a number of tentative com-
pounds for each of the matrix points. Factors used in the final
selection were:
The appearance of the compounds in the literature surveyed.
The availability of the compounds.
The variation of functionalities within the group selected.
A suspected or known carcinogenicity of the compounds or
similar compounds (but this was not a critical consideration)
13
-------
The aliphatic and aromatic compounds that were finally chosen are
given in Tables 4 and 5, respectively.
TASK 2 - SORBENT EVALUATION
Capacity Studies
The evaluation of the sorbent materials (Tenax-GC, Porapak R,
Porapak N, Ambersorb XE-340, and SKC activated charcoal) began
with studies of the abilities of these sorbents to quantitatively
trap various organic compounds. Capacities were determined for
as many of the 18 test matrix compounds as possible on each of
the 5 sorbents. Additionally, some low molecular weight ali-
phatics (methane, ethane, propane, pentane, and hexane) were
used to evaluate the carbonaceous sorbents, Ambersorb XE-340 and
SKC activated charcoal, for their abilities to trap extremely
volatile materials.
The experimental technique that was used to estimate sorbent ca-
pacities is a commonly cited, gas chromatographic technique in-
volving the generation of Arrhenius plots. This technique
involved packing chromatographic columns [0.9 m (3 ft) long, 0.6 cm
(1/4 in.) O.D., 2 mm I.D.] with known weights of each of the five
sorbent materials and conditioning these columns according to the
specifications of the sorbent manufacturer. A column was then
placed within a gas chromatograph and set at its highest allow-
able temperature under a flow of 30 m£/min of nitrogen. Samples
of the test matrix and/or low molecular weight aliphatic com-
pounds were then injected into the column and detected by a flame
ionization detector (FID). Sample sizes were within the Henry's
law region such that retention times were not affected by varia-
tions. The data obtained included the retention time (RT) which
was considered to be the amount of time from injection to peak
maximum and the first elution time (FET) which was the time from
injection to peak initiation. These points are indicated in
14
-------
TABLE 4. MATRIX OF ALIPHATIC TEST COMPOUNDS
Volatility
Hydrocarbons
Low
Low n-Hexadecane
(bp 205+°C)a
Medium iso-Octane
(bp 120-175°C)
High n-Butane
(bp <50°C)
Polarity
Halogenated compounds, Nitro compounds, nitriles,
aldehydes, ethers, amines, and strong acids,
ketones and esters alcohols and phosp_hates
Medium
Hexachloro-1 , 3-butadiene
(bp 215+°C)
Bis (2-chloroethyl) ether
(bp 135-185°C)
Propylene oxide
(bp <65°C)
High
Succinonitrile
(bp 225+°C)
Ethylene glycol
(bp 150-200°C)
Acrylonitrile
(bp <90°C)
aTemperatures in parentheses represent boiling point values used as criteria for
inclusion in the various volatility classes (low, medium, high).
TABLE 5. MATRIX OF AROMATIC TEST COMPOUNDS
Polarity
Hydrocarbons
Low
>
H
,-H
H
+J
(0
O
Low Phenanthrene
(bp 250+°C)a
Medium Naphthalene
(bp 200-245°C)
High Benzene
(bp <165°C)
Halogenated compounds,
aldehydes, ethers.
ketones and esters
Medium
4-Bromodiphenyl ether
(bp 255+°C)
1,2, 4-Trichlorobenzene
(bp 210-250°C)
Benzyl chloride
(bp <180°C)
Nitro compounds, nitriles,
amines, and strong acids,
alcohols and phosphates
High
Nitroaniline
(bp 280+°C)
Nitroanisole
(bp 230-260°C)
Phenol
(bp <210°C)
Temperatures in parentheses represent boiling point values used as criteria for
inclusion in the various volatility classes (low, medium, high).
15
-------
Vg
Figure 4. By multiplying the retention times and first elution
times by the flow rate and dividing by the amount of sorbent
(50% \ /
Vg^/ and FET I
volumetric capacities, respectively, were obtained. Sample
injections were then made at a variety of column temperatures to
obtain similar data.
An example of the experimental information obtained with Porapak
R is given in Table 6. These data were used to obtain an Arrhe-
nius plot (log Vg vs. 1/T where T = absolute temperature) .
Figures 5 and 6 are examples of these plots for the
Vg and
FFT "FFT1
Vg values, respectively, for Porapak R. The Vg values were
then extrapolated to 20°C to obtain the approximate ambient
(FFT1 \
Vg?Aor ) of this sorbent for the test
compounds. Table 7 gives this information for Porapak R.
RETENTION TIME
(RT)
INJECTION
FIRST ELUTION TIME
(FET)
Figure 4. Elution analysis method for
sorbent capacity determination.
TABLE 7. CAPACITIES FOR SELECTED TEST
COMPOUNDS ON PORAPAK R
FET
Compound name
Vg,
Vg
n-Butane
Propylene oxide
Acrylonitrile
iso-Octane
Benzene
4.93
12.4
16.8
3,550
175
16
-------
K
«
f^C
ft
«
O
53
O
Q
53
O
M-J
O
u
r .
U3
W
EH
Q
W
EH
U
w
h-H*
W
Cfi
OH"
0
UD
w
H
EH
H
U
f£
PH
i^
U
Q
H^
fQ
Cfl
W
s
M
EH
(-^
O
EH
S
W
EH
rv^
*
H
|_I]
m
25
t"^
a
3
1
§
U
a>
c
01
N
C
V
m
0)
c
(d
£
i
0
en
rl
OJ
rH
H
rl
-P
rl
C
0
^
O
<
Q)
a
H
X
o
ai
c
a)
a
o
0^
0)
c
id
3
m
c
1 CF1 Gn 'X x^
X^ X^ X^ X^ O1 D^ o< cij
oi o? PI r)
o n o o o o
OO OO OO rHrH
CrHCrH CrHCrH COCO C C
E X B X S X e X E E E S
xx o o
5 ^ N "! rH ". CO ^ CN ^D S ^D *°. S ^ rH
.(N'(N .^.CO .LT1.(N C0*m*
OCNOrH r-HLTlOrr, mrHCNJ,-! (NJ^fsJ^
II II II II II II II II II II II II II II II II
EHOIEHIJ, Ei&>EHCn BOiE-iOi EH DI EH d
* dP** £ EH^ * *^ § EH^" * *"" S EH"* ^ t*"* S E-I'*
OW OW OW OW
tP Cf tn XX
X, X^ (Jl X^ CA CO cai 0^
Ck? o^ 0$ O O
O O O O O
OO O OCOO rHrH
CrHCrH CO rH -HOCO
H -H -rl rH C E rH -H rH C X C X
E*S* S -H>< m E 'so 'so
u"Ji lO^r-"- cof*^ Lfi "*o in
* rH CT^ *CN»B> S?w> £>&:> £>u>
ow ow ow ow
mh intu mtu inw
XX XX XX tPCn
o o o o o o
oo oo oo oo
CrHCrH CrHCrH CH rH CO O
H -H -H -H -rl C -rl r-H C rH
EXEX EXEX BX-HX E -H
^r oiH icminm OTSr~ m^5**
rH.. .CM.kO -CO-CO .O-IO
OrHOLD OCNOrH rH'tJ'Or'l ^DrOLDfNl
II II II II II II II II II II II II II II II II
S>w> S^wt^ S>w> w^ws?
OW OH OW OW
int. men mtu mtu
xx xx xx tpcr
o o o o o o
oo oo oo oo
CrHCrH CrHCrH CrHCrH CO O
H -rH X
CNkOO*!T rO «rHC*J {D *f * ^fC^C^O
OCTiOCN OrHOt^- OrOCrH COrHCNrH
II II II II II II II II II II II II II II II II
SS'wS' oJS'w:? Sffw^ o:>'w>'
ow ow ow ow
tp t7> CP 01 CT1 tT1 D1
XX XX XX Oi X
d}
(0 O O O O O O
oo oo oo oo
CrHCrH CrHGrH CrHCrH CO rH
rl -rt -rl -rl -H -rl -rl rH C
SXEX EXBX SXEX E -rix
t~ tM * TT vD^O^f OMOin rHX^f
rHw> o;>w> oi>w> «>w>
dPWEn 4(>WEH aft fn ^ tf>b£H
ow ow ow ow
cs (M o o
cs co in o
CN H rH rH
17
-------
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
n-BUTANE
PROPYLENEOXIDE
ACRYLONITRILE
BENZENE
ISO-OCTANE
o
o
CSJ
4.0
3.0
1/T (K) x 10:
50%,
2.0
1.0
Figure 5. Plot of log JU'0Vg vs. 1/T for selected
test compounds on Porapak R.
18
-------
-3.0
-2.0 -
-1.0 -
0.0 -
C7>
1.0
2.0
q 3.0
4.0
5.0
6.0
7.0
8.0
n-BUTANE
_ PROPYLENEOXIDE /
ACRYLONITRILE
BENZENE
ISO-OCTANE
4.0
3.0
2.0
1.0
1/T (K) x 10
Figure 6. Plot of log FETVg vs. 1/T for selected
test compounds on Porapak R,
19
-------
The volumetric capacities for as many of the test matrix and low
molecular weight aliphatic compounds as possible were determined
for the other sorbent materials in a similar manner from the
plots shown in Figures 7, 8, 9, and 10, and are compiled in
Table 8.
The volumetric capacities (£/g) determined in these evaluations
estimate the volume of air that can be sampled per gram of sor-
bent materials before sample breakthrough (i.e., before exceeding
capacity). This was determined to be the most applicable measure
of capacity for our studies since we wish to sample quantitatively
(i.e., with no breakthrough).
Methods [4] are available for determining saturation (or weight)
capacities, but for this project it was felt that sampling tech-
niques, such as the use of backup tubes and varying sampling
volumes, would be an appropriate manner of obtaining valid
samples in the event of highly concentrated samples.
As it is a long and arduous task to determine the capacities of a
large number of compounds on several sorbent materials by the
method just described, it was thought that a means of correlating
sorbent capacities with compound properties would be highly
desirable. Such a means of correlation would permit the estima-
P'p'T1
tion of the Vg2f.0 of any compound on any sorbent by knowing
the value of the particular property or properties and how this
property is related to sorbent capacity. Then, if a change or
addition should occur in the list of compounds of interest, one
could state whether the new compound is probably, possibly, or
not likely to be quantitatively trapped by a particular sorbent
[4] Gallant, R. F., J. W. King, P. L. Lewis, and J. F. Piecewicz
Characterization of Sorbent Resins for Use in Environmental
Sampling. EPA-600/7-78-054, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1978.
151 pp.
20
-------
o
o
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
BUTANE <
PROPYLENE OXIDE
ACRYLONITRILE /
BENZENE A
ISO-OCTANE
BENZYL CHLORIDE
BIS(2-CHLOROETHYL)
ETHER
PHENOL
4.0 3.0 2.0
1/T (K) x 103
PET
Figure 7. Plot of log Vg vs. 1/T for
test compounds on Porapak N.
1.0
21
-------
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
D LOW POLARITY
A MEDIUM POLARITY
O HIGH POLARITY
©-
16-
v:
/
i/
i.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
n-HEXADECANE
HEXACHLORO-1.3-BUTADIENE
SUCCINONITRILE
ISO-OCTANE
BIS(2-CHLOROETHYL) ETHER
ETHYLENE 6LYCOL
n-BUTANE
PROPYLENE OXIDE
ACRYLONITRILE
PHENANTHRENE
4-BROMODIPHENYL ETHER
o-NITROANILINE
NAPHTHALENE
1,2,4-TRICHLOROBENZENE
m-NITROANISOLE
BENZENE
BENZYL CHLORIDE
PHENOL
4.0
Figure 8.
20° C
3.0
l/T(K)x 103
2.0
1.0
FETT
Plot of log Vg vs. 1/T for
all test compounds on Tenax-GC,
22
-------
o
o
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
METHANE
ETHANE
PROPANE
BUTANE
PENTANE
HEXANE
o
o
o
CO
I
4.0
Figure 9.
3.0
l/T(K)xlO
3
2.0
Plot of log Vg vs. 1/T for
standard gases on Ambersorb XE-340,
1.0
23
-------
en
>
UJ
O
O
-4.0
-3.0 h
-2.0 h
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
METHANE
ETHANE
PROPANE
BUTANE
PENTANE
4.0
Figure 10
3.0
1/T (K) x 103
FET,
2.0
l.C
Plot of log """Vg vs. 1/T for
standard gases on SKC charcoal.
24
-------
1
1
i
<3
c~*
fdj
Q
rv* »
w
EH
W
^
K
H
H i
H
«
' i
lJ
O
en
Q
§
s^
EH
H
u
f^
CM
_
O
6
H
EH
(^
t 1
H
CM
g
0
u
00
*H
EH
nj
0
CJ
in
.2
^
0
1
X
SI
o
U]
o
"g
2
*
*
s
rt
id
&
0
i
Q
O
1
C
01
*
Oi >
O EH
"o
So,
01 \
t?" °*
t,
Oi >
0 H
rH U
*
CJ
o
CM Ol
> =-?
EH
W
ft)
0-
0 EH
H pd
CD
"u
0
0
CM O!
Oi X
£,* **
&)
Oi
01 >
0 EH
^ £
_
G
O
CM 01
O! X
H
t)
£71 L>
O E"i
rH W
U
0
O
CN tji
t^ °*
w
p-l
1
u
5 rH
0
- -p
e -H
o
O ^0 r*' O co 'D
CN O o CN m m
I i
cj~i c~i in
O CJ m O
O \D
rH ^D
III 1 1
m CN in
1 1 1 S 1 0 CN I
O rH rH
n
^f CN \Q
III 1 rH rH 1
O r- rH
co ^r cn
0 00
i
ill i * i
^ggfNtngg^
^O^COCNCOrn^o
rHm^mr-Lninco
CN rH CT* m CO
cocococsommcN
I-H CN rr i m m m ^
OJ
-H O
X rH
C -H
,H
Q) C fi -H O
n3Cra-PC>irHX
0) 4J iH 1 1 SH O 1
1
1
1
CO
fff
CN
01
n
CN
^
s
CN
m
r-H
^r
<^
rH
CO
rH
CO
"
d
r^
rH
r^
,
O
rH
0
o
>1
CD
H
4->
(
' 1
1 1 1 i 1 1 1 1 1 1 1 1
1 1 1 1 i III 1 1 1
1
1 1 1 1 ll.lt 1 1 1
r- o ro
r- CN ^r
, t? CO ) xT 1 ) 1 1 1 1 i
CN m £,
rH
fO
r- tf) 10
o o o
rH rH rH
1 X X 1 XIII1 1 1 1
en \o CO
U> fN
1 1 1 1 l l l l t l l i
1(11 l i 1 1 1 1 1 1
mi^r-m mococNm m r- c^
Or*-Oin o^ocNrH "~* ^ o
^m^rro ^^^T^oin ^r vo xf
O 0 00000 O 0 0
r-t r-H rHi-Hr-HrHrH rH i-H rH
X X XXXXX X X X
OSSS 0°^rHG}r0. In ^. rH
. "d« CN CN rH rH . CN
rH^r l*rHi-H i-H i-H
in m
^HrH rHrHrHrHCN rH CN CN
OICN^T xfkD^DIf"- O vD !"-
-H rHrH rHr-lr-H ^H CN CN CN
\D tT> CN ["- iH]
t*~ O O xf i 1
IrHICN ^Tlir-CO 1 1 I
rH rH rH r ( r-H
r-t
XI >i 1
(IJtJ^OCiJ H >iro
rHHCJCr-H OC
H rH 0 -rHOOOirH 0) rHXl
n ooiq -ttncc^: -c ioj
4J HCO -H-HQJfOOXiC^ OC
H £OjrH CCMU-HOl-H iHCD
C O rH j£& fl (Tj ^ fl> i-4 C ^rQ 0 'H
Cr-HrHX:i OJiHiHCfOI NgOlXlfl
H 0 >i 4J CM j +J -P 2SSci)I!CCNX5CQCUXXl
3 XJ 0) <0 -H 1 1 X I I fl t
1
1
^
O
rH
OJ
rH
(U
-p
0)
e
H
0)
a
X
o
1 1J
CP
rcJ
rH 0>
rH P
H C
JJ CD
Q> 0
-P rH
O (Q
P. ffi
cfl XI
25
-------
under certain conditions. If sorbent materials were to be used
in series for sample collection, one would also have an idea on
which sorbent the majority of a certain compound is likely to
occur. The general advantage of a means of correlating sorbent
capacity and compound properties would be to eliminate the nec-
essity of experimentally determining the capacity of every com-
pound on every sorbent.
The forces that determine the interaction of an organic compound
with other materials (neglecting vapor-to-liquid phenomena of
nucleation and growth) are: E,, the dispersion or London forces;
E , the permanent dipole/dipole forces; and E, , the hydrogen-
bonding forces. These forces are additive and the total inter-
action energy may be expressed as E = Ej + E + E,. Dividing
both sides by the nuclear volume of the material, one obtains:
E, + E + E,
E/V = -£ ^ (l)
since
E 2
6=1 (2)
where 6 = the solubility parameter
then
6 = 6 ,2 + 6 2 + 6, 2 (3)
d p h
The solubility parameter has been well described in the litera-
ture [5] .
[5] Burrell, H., and B. Immergut. Solubility Parameter Values.
In: Polymer Handbook, J. Brandup and E. H. Immergut, eds.
Interscience Publishers, New York, New York, 1966.
pp. IV-341 - IV-368.
26
-------
Many experimental methods exist to determine 6 for a compound, and
many more methods exist for estimating 6 from compound properties
such as surface tension. A very useful method of estimating 6
from the structural formula, density, and molecular weight of a
compound has been developed by Small [6].
Accordingly, 6,, 6 , 6, , and 6 were examined for a number of com-
pounds. However, no correlation of the separate or total forces
TTTTT"
with the V920°C was f°unc^- One would expect a correlation
with the dominant force factor, if one should exist, and a corre-
lation with 6 would seem even more reasonable. However, of two
compounds with similar 6 values, the compound with the higher
molecular weight (or size) would be expected to have a larger
FET
V^2n°r on a 9i-ven sorbent due to its lesser volatility.
Therefore, various molecular weight adjustments of 6 were tried.
The simple expedient of multiplying 6 by the molecular weight
gives a relative size dependent factor, 6=6 MW, which has
1 / m
the units of (g2*cal) 2 / (cm3-mole) . The molecular-weight-
modified solubility parameters (6 ) for the test matrix and low
molecular weight aliphatic compounds, both from literature and
calculated [4] solubility parameter values, are given in Table 8
along with sorbent capacity data. The data for 6 are plotted
FET m
versus the log Vg~nOp for SKC charcoal, Ambersorb XE-340,
Porapak N, and Tenax-GC in Figure 11 and for Porapaks R and N in
Figure 12. The correlation coefficients for the "best" lines in
these figures are given in Table 9.
Note that the most divergent results occur for a few halogenated
compounds on Tenax-GC. Unusual behavior of such compounds on
Tenax-GC had been noted previously during capacity studies. It
was postulated that some sort of solubility effect was possibly
occuring since Tenax-GC has a solubility parameter similar to
[6] Small, P. A. Some Factors Affecting the Solubility of
Polymers. Journal of Applied Chemistry, 3(2):71-70, 1953
27
-------
LPl
o
ro"
o o
^-i CD
C3
O-'
CD
od
CD
-O
CD
CD
CD'
CM
I
6/7 '
u
O
O
CM
tn
>
a)
PM
&l -P
O C
iH QJ
T3 M
C O
(0 tti
e c
-------
cc.
o
O
ct:
< -i
> 0)
EH +J
W (0
fc e
CP -P
O C
rH CD
X3
T3 5-1
C O
tO tn
ec
-------
that of halogenated compounds and is known to be soluble in such
halogenated solvents as chloroform, carbon tetrachloride, and
methylene chloride.
TABLE 9. CORRELATION COEFFICIENTS FOR THE "BEST" LINES OF
"PT^T1
6m VERSUS LOG Vg2QOC PLOTS (FIGURES 11 AND 12)
Correlation
Sorbent material coefficient
SKC charcoal 0.996
Ambersorb XE-340 0.999
Porapak N 0.993
Porapak R 0.944
Tenax-GC 0.806
Tenax-GC (excluding
halogenated compounds) 0.916
Desorption Properties
A number of experiments were also conducted to determine how
amenable the selected sorbents are to thermal desorption. These
experiments were of a preliminary nature since the analytical
system to be used for sample analyses was not completely devel-
oped. Also, the desorption properties of specific compounds for
analysis depend upon the compound as well as the sorbent mate-
rial, and the test matrix compounds are not necessarily the
compounds that will be of interest in future field studies. The
preliminary experiments that were conducted attempted to evaluate
the sorbents for efficient compound desorption and extraneous
background due to contamination and/or sorbent characteristics.
Further experiments with the carcinogenic compounds of interest
for field sampling will be conducted as part of the analytical
development of Phase II.
For the desorption efficiency experiments, 15 cm (6 in.) long,
0.6 cm (1/4 in.) O.D., 4 mm I.D. glass sampling tubes were packed
30
-------
with the sorbent materials listed in Table 10 and evaluated on
the desorption/analytical system depicted in Figure 13. Note
that the Chromalytics Concentrator oven contains a six-port, two-
position valve, which is sketched in Figure 14. This valve
offers the alternative of direct syringe injection of a standard
or thermal desorption of a sampling tube. Standards were pre-
pared with a 1,000 yg/m£ concentration in acetone and stored in
septum-capped vials. Chromatographic conditions were established
by direct injections of standards using a 2.1 m (7 ft) long,
0.6 cm (1/4 in.) O.D., 2 mm I.D. glass analytical column contain-
ing 1.125 g of Tenax, which was conditioned at 320°C overnight.
For statistical determination, replicate, direct injections
(^1 y&) were made of each standard under the chosen chromato-
graphic conditions and the areas of the peaks integrated by a
Hewlett-Packard Model 3354 Lab Automation System. Sampling tubes
were placed in the desorption chamber, conditioned briefly
(^30 min), and quickly investigated for background under the
desorption/chromatographic conditions to be used. Injections
(^1 y£) of the prepared standards were made onto the far ends
(i.e., farthest from the analytical column) of the sampling
tubes, with subsequent thermal desorption into the analytical
column for chromatography and peak integration. Desorption effi-
ciencies were determined in the following manner:
Desorption _ (Area of peak from sampling tube desorption)
Efficiency (Average areas/y£ of peaks from standard injections)
x 100
[Amount (y£) injected onto sampling tube] (4)
In order to establish a median and a range, the desorption effi-
ciency was determined at least three times for each compound on
each sorbent.
Since these experiments were preliminary only a few of the test
matrix compounds were used in the evaluation of the desorption
31
-------
*^"**
CO
o
D
1-3
A
,
O
o
IS
CO
CO
o
w
>
rf^
S3
CO
w
m
D
EH
Q
w
1-3
I-H
H
fa
1
EH
W
ffl
PH
0
CO
n
i^l
rfj
>^x
£
EH
f^
Q
W
PQ
D
EH
0
2
H
I-H
&J
^J
^
CO
0
W
03
"^
EH
0)
C )H
O 3
H 4J
4J »
JJ (0 4J
O M -H U
Id 0> g o
«*H ft -H
3 & rH
C 0) ^
fO -l-J
s
M-l
0 -P -H
C nj
4-> 0) -H
C X! in Cn
3 M a)
< g
C
0
H
4J
i
4-> O
0) MH
rH C
X
O C
01 O
4J
o
(d
4J
X
0)
0)
A O
H Z
r-|
id
^4
0)
4-1
10
e
4J
q
0)
g
0
o o
o o
n ro
m m
CN Cvl
o-> ro
o o
o o
^f ^
oo r-
rH O
rs m
rH i-H
O 0
g
0)
.c
4-)
-H
^
4J
x;
en
-H
c
q o
O r*
2 O
GO r^-
rn
* ^ . »
x: x;
ui en
0) 0)
& e
0 0
CO CO
X X
0 O
^0 vC
- v *
X X
rd id
q q
0) 0)
H H
O
O
ro
in
CO
O
O
*d*
Cft
CT^
ro
H
O
0)
q
4J
q
0
X5 ft
u
(0 'O
a) q
CD
\
J^-
vD
-
X
id
q
0)
EH
ooooooooooo
t^r~r-r~ooooooo
-i-HrHt irororororororo
0 0
(N Ol
'O'Dininmoooomin
(0 (BrHrHrorOrororororo
m m
r- r^
rH rH
O O O O
in in ^ c^» i i i i i i i
(N N iH iH
ooncovor^Oi-nror-o
kD CO f^ rO rH rH in i 1 CN rH "1
(N(Nroro^o>oooooo>
kOinOO^fr-l^i lroS ^ ^S
(0 (Tj (TJ
P£ W S IZ ^ t> ^ ,Q Q Q Q
H -H -H in JH *-i M
'rrf'fd'iS'rrtUUUW 88$
UtQ^QiCurC 03 (T3M i-i i^Vn
fC'O'Xf'O flJQJGJQJ
^ V^ ^ Wj (_l !_> C l Q Q Q Q
O O 0 Or*3J^t^£?GiE5G
(Xipu04dicoc/)co
-------
<
5
Or!
LU
T:
i
g
o
Q
U-
O
i
z
o
o
o n
o o
o a:
co
CJ
<
o
ce
a:
o
c
^E
<»»
E
o
CO
CO
<
o
on
CJ
s
o '
o
cc:
CJ
o
C5
^,1
\r\
<^i
1
on
O
a.
z
o
t
o
^
z
o
o
z
LU
>
o
to
CJ
^~
^
s
o
o;
~1~
CJ
-^
o
H-
CJ
LU
o
CJ
0
o
E
-0
o
o"
z
o
_J
E
CNJ
0
z
t
z
o
o
LU
o
CJ
o
2
CO
s
x
i
H-
LL.
o
LTV
CSI
1 t
1
g
;}
o
o
o
o
o
LT\
f-\
o:
o
)
CJ
£
LU
Q
z
o
L_
M
o
LU
s
^
Li-
en
fy
<
^
o
CC
UJ
h-
Q,
^^
o
CJ
0 ©
©©
ATOGRAPt
5
o
Cf.
3:
o
CO
o
VLETT-PACKARD
^^~
LU
^
^
r^
LTV
Q£
UJ
Q
C£
8
UJ
ex
h-
Ql
^
z:
o
VLETT-PACKARD
UJ
^
CVJ
r 1
1
§
( 1
g
LU
1
O
Q£
1
O
CJ
.YTICS THERMAL
TRATOR SYSTEM
is
S o
0 2
tXL O
3: o
o
HAMBER
o
o ^
1 LU
Q_ '
a: co
o *"
10 CO
.YTICS TUBEDE!
CONCENTRATOR
(
S§
0
o: o
X |
0
ESORPTIOI
O Of.
°3
£g
1
og
OUPLECONNECTI
iR TO THERMAL C
8|
J^ i
rH
nJ
C
O
-H
O
w
Q)
Q
0)
33
-------
SORBENT TUBE IN TUBE FURNACE
SWITCH IN "TRAP"
POSITION
(DIRECT) INJECTION PORT
CURRENTLY UNUSED (CAPPED)
CARRIER GAS SOURCE TOGC
DIRECT INJECTION OF STANDARDS INTO GC
SWITCH IN
"BACKFLUSH"
POSITION
SORBENT TUBE IN TUBE FURNACE
(DIRECT) INJECTION PORT
CURRENTLY UNUSED (CAPPED)
CARRIER GAS SOURCE TOGC
THERMAL DESORPTION OF TUBES INTO GC
Figure 14. Diagram of six-port, two-position
valve within the Chromalytics oven,
34
-------
properties of the sorbents. However, by relating desorption
efficiency to retention time data found previously (for capacity
studies using 0.9 m long, 2 mm I.D., 0.6 cm O.D. glass columns
filled with the sorbents) at or near desorption temperatures, an
idea of the types of compounds that are readily thermally de-
sorbed from the selected sorbents may be obtained. Table 11
gives the chromatographic conditions and raw data for the deter-
minations with the porous polymer type sorbent materials, while
Table 12 presents retention time data along with experimentally
determined desorption efficiencies. Note that several values are
above 100% efficiency. This may be due to either background
contributions to peak areas (especially for Porapaks), or slight
differences in peak shapes between direct injections and tube
desorptions resulting in different peak areas. In relationship
to the indicated retention times, the desorption efficiency data
suggest potentially easy quantitative desorption of compounds
with retention times (Table 12) of less than a few (^15) minutes.
This, in turn, indicates that Tenax is probably very amenable to
the quantitative thermal desorption of all 18 test compounds
(assuming no decomposition problems) and potentially of other,
less volatile substances. Compounds with retention times
(Table 12) of greater than about 15 minutes, such as benzyl
chloride on Porapak N, may also be amenable to quantitative
thermal desorption if the desorption period is prolonged and the
sample is reconcentrated in a capillary trap or on the head of
the analytical column.
Unfortunately, this experimental procedure, which worked quite
adequately for the porous polymers, proved not to be amenable to
the evaluation of desorption efficiencies of the carbonaceous
adsorbents. Much broader peaks resulted from the thermal desorp-
tion of sampling tubes compared with those peaks produced by
standard injection. This was due to the longer retention times
for thermally desorbed samples that resulted from their having to
pass through the sorbent tube as well as the analytical column.
35
-------
w
ffi
EH
DH
O
fa
C/3
rfj W
EH H
4 U
P 2
M
15 H
(^ r)
PH r i
fa
Q fa
^7 M
^-4 t^i
'"^
r'H
en O
O EH
H CM
*» _
1 1 O
p to
S3 K
O P
u
fa
u o
H
04 O
PH EH
o <;
o s
E~^ ^~f
j£ ^g*l
^j PH
o w
Pi EH
ffi W
U P
*
r '
rH
W
H" i
W
^^
£H
c
o
H
jj
a
o
to
01
Q
4J in
E 0
01
01
Q
.
O
C
0
0
co
01
(d
01
CP
fO
H
0)
>
<
01
CP u-i
id o
01 E-"
5* K
<
£
O
C
01
o
H
14H
UH
OJ
- 0)
Cd 4J
oj c
3 §
EH -rl
« g
fc
TJ °J
4J 3-
01
rH
fd
£
^
a u
g o
01
^
a
4-1
CP
c
H 01
rH £>
a 9
g 4J
tO
!/l
fc
"2
ffl o*
O 3
C ~s.
r8 01
4J JJ
3
iH O
0 0
UH
,
fJl
t3
iH
id C
id
4-1
(n
CJ
H
X
a oj
rd C
IH O
CP rl
0 4J
id 13
S C
0 O
i4 U
.C
u
0
c
3
a
o
u
r-^^1 CN m o ro rs co
CTlrHO rH O O rH rH rH
IT) C3^ 'tf in CO U"> CN v£ 01
OCNCO r-» oo o i£ o in
CQ o~i o^ OCNCO o cy» r--
\£ O r- *H rH O (JiCOCO
i-l CN rH i-t i-H rH iH (U
a si o u x
c 4J a 4J
0) U H E
H o g C 0) T3
3O rl 4J rH
4-> m o g 01
fd rH \ (_) £ C
iH g O 0 rl
01 O IH 0 O TJ E
a in o CN in c
g IJH UH ro CN id CO
a
EH
OJ
C
rO
14J
U
O
1
O
Ul
-1
CM en en o^
r~ en 10 VD
CTi ^0 VD CN
ro in <3* TJ*
co in "^ r^
kO in ro CN
ro ro ro ro
CO CO CO CO
o o o o
0000
rH H O O
rH CN ro *y
O
0
ro
O
O
S.
X
fd
C
OJ
CO
o
;o
CO
in
CO
o
rH
T3
01 CJ
go T3
g O rH
id in 01
V4 CN X
CP
O O T3
a fd
u
0) o c
^ O -rl
3 in g
4-> rH \
id cj
iH g 0
01 O CO
a M
g UH CO;
0)
£-t
01
c
01
r-l
rd
£
4J
r;
a
rd
2
3 XI
1 1
t |
0 0
ro m
A A
0^ \&
H rH
iH rH
rH rH
O O
CN in
CN rH
m m
ro r**
CN rH
*& ^2
id in cd
a rH a
id *= id
0 0
Cu PH
en ^"
CN rH
r- co
O rH
in t
CO ro
ro o
ro ro
rH rH
rH rH
1
01
a
g
OJ q
4-1 -H
g
rH
id T
q
H ^i
14H O
0)
X! 01
4J iH
rj
4-1 4-1
cd fd
rH in o
rH O rH
rH rH rH
ro vD ro
CO CM li)
rH en *r
CN Cn rH
CO ro CO
rH rH rH
in rH us
co o en
r^ co r~
o in o
CO kD CO
000
^1 CN ro
O
O
CO
0
0
ro
r-
3*=
X
fd
q
01
EH
CO
in
in
o
CN
CM
r-
P.'
t)
0) O -o
g O rH
g O 01
fd o x;
SH ro
tp fuj
o o c
^ 4J id
a
u c
0) 0 -rH
IH 0 g
3 0 X
4J CN CJ
to o q
iH g CN -H
01 O ro E
a M
E *4H CD, <3-
01
01
C
0)
_r^
4J
e.
01
xi
(X
in oo
en CM
^D CP
en ^*
ro ^*
O CN
P~ rH
CN ro
in rH
IN CN
01 CP
0 0
O rH
rH rH
O
rH CN
O
0
CO
O
o
ro
p.
X
id
q
QJ
EH
CJ
CO
CM
0
rH
en
CN
O
0
0
en'
01 CJ
§o
0
rd m
^4 CN
CP
o o
^ 4J
a
o
01 o
r. 0
3 O
4-1 CN
id
01 O
a M
g "H
OJ
EH
CM O
CP. O
1-1
CO CO
cp en
i£ CN
CO CN
ro CN
CO CO
rH O
rH rH
CP CM
in o
rH rH
rH rH
O
ro *r
CJ
o
CN
CN
O1
CN
O
in
en
CO
O
r-l
o>
si q
H
tJ g
c
rd T
C V-4
g £
\
U CJ
O 0
T O
in
C21 CN
OJ
c
m
u
01
a
id
X
01
i
q
01
4J
q
11
E
^
0)
a
QJ
j^
QJ
j_)
o
rH
rH
to
^i
o
MH
C
01
o
14
4J
-H
q
01
rO
3
Vj
01
-H
^
id
U
fd
01
C
o
H
H
§
fS
4-1
c
H
Vj
a
X
01
4J
id
c
H
E
o
ro
^i
U
4->
UH
id
c
o
H
U
a
OJ
T3
01
rH
a
E
rS
01
14H
O
C
O
H
4J
id
o
rl
"S
rl
0
Z
a
o
MH
01
0)
3
iH
CO
^
4J
C
OJ
11
r^
CM
C T3
H U
01 01
3 3
IN 01
tJ
C <1S
O T)
rj
rH CO
4J
01 01
rH
CO 01
H s:
iH -P
E O
iH
""' *S
>i c
(d 'rl
4J
C O
11
M OJ
(U i4
UH 11
H
3-
co x.
01
q 4J
o q
3
o o
OJ 0
4-1
0 T)
3 c
frj cd
q
O 01
0 g
H
01 4J
01 q
S 0
H
^ 4-1
q
d 0)
C 4-1
fd 0)
^4
CO
01
01 rp
rH fd
rd ^4
H 01
iH >
EH rO
U
36
-------
TABLE 12. RETENTION TIME DATA AND DESORPTION EFFICIENCIES AT
OR NEAR DESORPTION TEMPERATURE FOR SORBENT MATERIALS
Test compound
n-Butane
Propylene oxide
Acrylonitrile
Benzene
iso-Octane
Ethylene glycol
Succinonitrile
Phenol
Benzyl chloride
Naphthalene
Bis<2-chloroethyl) ether
Nitroaniline
Nitroanisole
Phenanthrene
n-Hexadecane
1,2, 4-Trichlorobenzene
4-Bromodiphenyl ether
Hexachloro-1 , 3-butadiene
Retention time, min
, Porapak N Porapak R
m @ 150°C @ 150°C @ 222°C
395 0.51 0.50 0.17
534 1.03 0.68 0.20
557 1.93 1.00 0.24
719 4.48 3.23 0.47
788 12.1 10.33 0.85
(113±l%)a (108±4%) (>132%)
906 -
1,073 -
1,176 -
1,253 110
1,269 -
c c
1,402 165
1,657 -
1,684 -
1,747 -
1,811 -
2,014 -
2,616 -
2,783 -
Tenax
@ 302°C
0.10
0.11
0.08
0.12
0.11
(106110%)
0.11
0.17
0.17
0.16
0.22
(95±2%)
0.15
0.41
0.33
1.02
(109±4%)
0.21
(98±3%)
0.20
0.61
0.18
Numbers in parentheses are desorption efficiencies.
Significant contribution from sorbent background.
*
'No indication of sample desorption after 30 min at experimental
conditions.
37
-------
The difference in the peak shapes made it impossible to accur-
ately compare peak areas with the standards by this technique.
For example, n-butane has a retention time of 0.6 min on a 0.9 m
column containing 1.11 g of Ambersorb XE-340 at 300°C, but desorp-
tion at the same temperature from a sampling tube containing
^0.6 g of this adsorbent resulted in a retention time of 1.4 min
and an average desorption efficiency of only 76%. Similar results
were obtained for n-butane with charcoal-filled sampling tubes.
It is uncertain how amenable these carbonaceous sorbents will be
to thermal desorption. It is possible that with the sample
reconcentrated after a prolonged desorption period, some of the
more volatile compounds could be analyzed. However, solvent
desorption of these materials may prove to be more desirable.
In addition to desorption efficiencies and retention time data at
or near desorption temperature, Table 12 also gives 6 values for
the compounds listed. This table indicates a relationship
between these retention times and 6 , especially for Porapaks R
and N. This relationship suggests that 6 may be used to deter-
mine which compounds will be quantitatively desorbed from sorbent
materials.
Since a fair correlation has been demonstrated for 6m and sorbent
capacity, and a relationship potentially exists for 6 and de-
sorption efficiency, then 6 may be useful in determining which
compounds are likely to be quantitatively adsorbed on and subse-
quently desorbed from a particular sorbent material. A qualita-
tive comparison of 6 with sorbent capacity and desorption effi-
ciency is given in Table 13. The entries in this table are
defined in the key beneath it and are based upon a combination of
experimental data, demonstrated correlations, and chemical
intuition. The enclosed regions indicate compounds which are
probably or possibly quantitatively trapped and are probably or
possibly quantitatively desorbed from the sorbent materials indi-
cated. These regions correspond to ranges of <5 values which are
listed in Table 14.
38
-------
EH
H
rH1
H
CO EH
W D
H
EH EH
A X
W W
ft ffl
O P4
CM O
ft CO
B rH
& O
W
OQ CO
Pi W
O O
CO S
O «
IS
EH W
s w
O EH
H W
EH P
D O
3*
> e
W co
w o
> EH
H
EH P
H <
(-5 ft
< g
D O
oa
ro
W
rH
rH
0
u
^
(0
x;
tj
u
t/}
X!
rH
0
U
0)
XI
1
2
_y
. r*- r><
* i
N.^~%
(V. 0. f>
%.>>.%.
o o o
O 0 0
o o o
O 0 O
000
o o o
en n
rH r-*
in i in
in *» r~
o m in
in in in
CD
T3
H CD
X rH
O -H
SH
0) (1) 4J
C G -H
rO 0) C
-P rH O
G !>lrH
01 a >
CM O SH
1 M U
C CM <
X X
^~-i
c\. r>
>^^-
0 0
o o
0 0
0 O
0 0
0 O
n rH
VD r-
en c^
CN rH
us r~
0)
C 0)
tO C
X 0)
CD N
33 G
1 (U
G ca
x x
N."^.
v%.
%.%.
N."^
0 O
o o
n
CO 1
CO ID
00 O
r- c*
rH
O
0
rH
CD tji
G
tfl l
o xi
01 -P
H H
xxxx xxxxx x x x
~~>^"^N. %. ^ >r. ^ %. %. ^V "^.
XXXX XX XX X X X
~^.%.^a.%. N. %. ^v N. "^ ~~>. "S. "V
r. c^' r. r>-
XXXX X X X
^Ifc "^fc ^ ^V ^> ^* ^X ^ ^ ~^>. ^V ^».
r>. ^ p. r-
XXXX X X X
^.^^V^ ^^.^^.^i >» ^>. %.
N^^v^x N.^^."^."^ >» %. N.
rf^ rn Q (N r°^ ^j* ^D ^* \& ro
r~- uirH oinco VD rH rH co
OICN^T wvo^oir- o vo t-
rH HrH rHiHrH rH CN CN CN
VD CTl CN r- rH
r~ VD O -^ rH
IrHICN « 1 1 r» 00 1 1 1
rH rH rH rH rH
r7
0) XI O rH 1
tDTJ-POIl) M>in
rH-HCDCrH OG
H SH O -HOCDCDrH CD rH
SH OtDSH rHWGGx; X! ICD
4J rHGO -H-HCDi-pcNx;-p+jiox'rGox;u-P
UGNXI-^-P-H-HCO) >CDSH-PtflD
Q QJ C Cu 01 QJ 2 12 Or) i-C (N .Q PQ QJ i^ J3
pXittJlO-H 11X11- 1 0)
UlCuCQZOJ OSCMGrH <» K
.(
CD
rH
XI
ifl
-p
o
4J
^
CD
X
o
HO°
o x
to
in
O
CM
O
M
CM
>< CD
rH >
0) -H
> -P
-H rfl
-P -P
10 -H
-P 4->
H C
-P fO
c a
rd a
OX!
Cb O
ra to
M CD
EH Q
39
-------
TABLE 14. RANGES OF UTILITY OP SOLID
SORBENTS BASED ON 6
Sorbent
SKC charcoal
Ambersorb
Porapak N
Porapak R
Tenax-GC
6m (Trap)
^350
^450
V750
^750
^950
to
to
to
to
to
to
6 (Desorb)
^600
^750
^1,500
^1,500
>2,800
Based on these results a sorbent material may be used to sample
most compounds with 6 values falling within its &m range. Al-
ternatively, a sorbent material may be selected for sampling
studies with a particular compound which has a 6 value within
the 6 range of utility of the sorbent. For example, a compound
with a 6 value of 2,000 should probably be sampled using Tenax-
GC. A compound with a 6 value of 1,400 falls within the ranges
of utility of Porapak N, Porapak R, and Tenax-GC. Since 1,400
is close to the 1,500 6 desorption limit of the Porapaks, Tenax-
GC would again be the best sorbent (of these five) to evaluate
first for sampling this compound. However, a compound with a 6
value of 1,000 would probably be easier to sample using one of
the Porapaks since this value is near the 6 trapping limit of
Tenax-GC. A problem potentially exists for compounds with &m
values around 750, since this is the
-------
The other area of interest during the evaluation of the thermal
desorption properties of the selected sorbent materials was
potential extraneous background resulting from contamination
and/or sorbent characteristics. Sorbent backgrounds can be com-
posed of a variety of contaminants from a number of different
sources. Depending upon their degree and nature, these contami-
nants can lead to complications during both sample collection and
analysis. In order to minimize problems from background contami-
nation, it is best to know as much as possible about what types
of contamination can arise, how they usually occur, and how best
to minimize all sources of contamination. In this way, complica-
tions, such as in situ reactions and interference during analy-
sis, can be minimized and possibly avoided.
Generally, all commercial sorbent materials as received from sup-
pliers contain significant levels of background contamination as
the result of production, packaging, and/or transportation. Al-
though most of this initial contamination may be removed by sol-
vent (before packing in tubes) and/or thermal (usually after
packing in tubes) conditioning, such conditioning may also
introduce contaminants. Backgrounds may be introduced from trace
impurities in the solvent and the gas stream, and also from
decomposition products of the polymer. Obviously, one does not
heat a polymer above its temperature limit or extract it in a
solubilizing solvent, but some polymeric breakdown can occur
under other situations as well. For example, Chromosorb 102 is
reported to be stable to 250°C by its manufacturer. However, the
experimental data shown in Figures 15 and 16 suggest three possi-
ble regions of varying thermal stability [7]. Below 170°C both
differential scanning colorimetry (DSC) and thermal gravimetric
[7] Adams, J., K. Menzies, and P. Levins. Selection and Evalua-
tion of Sorbent Resins for the Collection of Organic Com-
pounds. EPA-600/7-77-044, U.S. Environmental Protection
.Agency, Research Triangle Park, North Carolina, 1977.
61 pp.
41
-------
o
O
X
f
o
60°C'
§j 0 50 100 150 200 250 300 350 400 450 500
§ TEMPERATURE, °C (CORRECTED FOR CHROMEL ALUMEL THERMOCOUPIES)
Figure 15. Differential scanning calorimetry data
for Chromosorb 102 [7].
10
en
o
E 4
0
"60 V
0 50 100 150 200 250 300 350 400 450 500
TEMPERATURE, °C (CORRECTED FOR CHROMEL ALUMEL THERMOCOUPLES)
Figure 16. Thermogravimetric analysis data
for Chromosorb 102 [7].
42
-------
analysis (TGA) data indicate a region of excellent thermal stabil-
ity. Between 170°C and 250°C the DSC data indicate a region of
change in thermal properties, perhaps relating to a phase change
in the polymer. A change is not indicated by the TGA data until
^245°C. Above 250°C both the DSC and TGA data definitely indicate
the thermal breakdown of Chromosorb 102. Conditioning this
sorbent at 240°C might result in a lower background at a 200°C
desorption temperature since more of the original background
would be removed. However, it could produce a higher background
at 200°C due to the increased number of thermal decomposition
products that would be produced if the inflection in the DSC
traced at ^200°C indicates the onset of thermal decomposition.
Each sorbent will have thermal properties that are unique and,
therefore, conditioning procedures should be validated and opti-
mized for each sorbent material. As a precaution, one should
also be aware of the types of compounds each sorbent produces as
the result of thermal and/or chemical degradation. Table 15
lists several of the major thermal decomposition products that
could be expected from the selected sorbent materials, as well as
the chemical compositions and temperature limits of the sorbents.
Since the temperature limit of Ambersorb XE-340 was not stated by
its manufacturer, a freshly packed 0.9 m long, 0.6 cm O.D.,
2 mm I.D. glass column of this sorbent was prepared, conditioned,
and analyzed by GC/MS to verify that thermal decomposition was
not occurring at the conditioning temperature of 350°C. This
column was placed in the GC oven of a Hewlett-Packard Model 5982A
GC/MS system and subjected to stepwise temperature increases
(50°C intervals) up to 250°C. The effluent from the column was
continuously monitored by mass spectrometry (MS) and showed none
of the expected thermal decomposition products of the styrene-
divinyl benzene polymer. Indeed, the only effluents observed
were COa, SC-2, argon, and traces of a few low molecular weight
aliphatics. Therefore, thermal decomposition of Ambersorb XE-340
43
-------
does not occur at 350°C, which is above the anticipated operating
temperatures.
TABLE 15. EXPECTED THERMAL DECOMPOSITION BEHAVIOR
OF SELECTED SORBENT MATERIALS
Sorbent
Chemical composition
Major thermal
Temperature decomposition
limit, °C products
Porapak N
Porapak R
Tenax-GC
N--Vinyl pyrrolidone
N-Vinyl pyrrolidone
190
250
2,6-Diphenyl-p-phenylene oxide 400
Vinyl pyrrolidone
Pyrrolidone
Pyrrilidiene
Vinyl pyrrolidone
Pyrrolidone
Pyrrilidiene
Alkyl benzenes
Styrene
Benzene
Alkyl phenols
Ambersorb XE-340 Carbonized styrene-divinyl >350
benzene
SKC activated Carbonized organics >400
charcoal
Major decomposition products not determined.
Once the conditioning procedure for a particular sorbent has been
established, care must be taken to prevent additional contamina-
tion during subsequent handling. The handling and storing of
sampling tubes should be optimized without being unreasonably
complicated or expensive. The techniques developed should also
ensure the integrity of all samples collected (i.e., no sample
loss).
Finally, backgrounds from other sources in the desorption/analyt-
ical system should be eliminated or reduced to the greatest
extent possible. These can arise not only from the usual sources
44
-------
of chromatographic background contamination (e.g., column and
septum bleed) but also from sources that are unique to the method
of desorption and sample introduction. These less obvious diffi-
culties could include dust being introduced into the desorption
chamber or fingerprints on the sample tube, depending upon the
desorption chamber design. Generally, background contamination
introduced by a desorption/analytical system is best minimized by
using forethought when designing the system.
A series of experiments was conducted to evaluate the background
levels of the materials being studied and possible sources of
background within the desorption/analytical system. Sampling
tube backgrounds were evaluated by using the desorption/analyt-
ical system sketched in Figure 17 and the sampling tubes listed
previously in Table 10. Examples of two experimental background
chromatograms for a Tenax-GC tube are depicted in Figure 18.
Below are some of the findings determined during these evalua-
tions of sorbent backgrounds.
Conditioning
No difference was noted between solvent-extracted and
non-solvent extracted sorbents (for the sorbents studied
and the experimental system used).
For Porapak R, conditioning at 240°C produced less
background upon desorbing at 170°C than with condi-
tioning at 175°C and desorbing at 170°C.
Thermal conditioning removed a great deal of contami-
nation, most within the first hour.
After conditioning, the backgrounds of Porapak R and N
tubes were above the background of a blank control tube,
while the backgrounds of Tenax-GC, Ambersorb XE-340,
45
-------
V V \ V
c
4
3 C
X;
=>
«£i
c=>
x"
X
X
O
O
( >
I
X
X
X
X
O O
O O
o
O
0
o
o
/
o
1 1
1 1
11
a
:
n
n ifli
\9/
(D
©
5710A HEWLETT-PACKARD GAS CHROMATOGRAPH
7132A HEWLETT-PACKARD CHART RECORDER
CHROMALYTICS THERMAL CONTROLLER TO 1047
CONCENTRATOR SYSTEM
CHROMALYTICS TUBE DESORPTION CHAMBER TO
1047 CONCENTRATOR SYSTEM
THERMOCOUPLE CONNECTION FROM DESORPTION
CHAMBER TO THERMAL CONTROLLER
DESORPTION CHAMBER ENTRANCE AND (0-RING) SEAL
NITROGEN CARRIER GAS (30 mllmln)
DESORPTION CHAMBER/GC INJECTION PORT (300 °C)
CONNECTION
GC OVEN (320 °C) CONTAINING EMPTY 0.9 m LONG, 0.6 cm 0. D.,
2mm I.D. GLASS COLUMN
FLAME IONIZATION DETECTOR (350 °C)
GC CABLE TO CENTRAL 3554 HEWLETT-PACKARD
COMPUTER SYSTEM
Figure 17.
Desorption/analytical system used
in tube background evaluations.
46
-------
a:
<
o
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0
RANGE =10
ATTENTUATION :64
INITIAL BACKGROUND OF A TENAX TUBE
375 ppm COR~110 ng C/s
\
\
10 ppm COR~3ng C/s
\
. \
V
N._.
..* v
* 1 1 1 1 1 1 1
BACKGROUND OF TENAX
/TUBE AFTER CONDITIONING
/
***« «
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
TIME, min
Figure 18. Experimental chromatograms of the backgrounds of a
Tenax-GC sampling tube before and after conditioning,
47
-------
and SKC activated charcoal matched the background level
of the blank control tube.
Handling and Storage
A desorption system designed to eliminate the exterior
of the sampling tube from analytical desorption greatly
diminishes problems with handling and storage.
Capping tubes with metal fittings and storing them in cul-
ture tubes with Teflon-lined caps seems the best method
of tube storage to eliminate extraneous contamination.
Glass sampling tubes offer a more inert surface than
stainless steel with fewer contamination and reactivity
problems.
Evaluations of Sampling Tube Pressure Drops and Design
One limitation that automatically applies to any portable air
sampling system is the ability of a portable pump to pull air
through sorbent-filled sampling tubes at a desired rate. This
ability is directly related to the pressure drop across the sam-
pling tube. For a particular pump to pull air through a sampling
tube system at a specific rate (0.5 to 3 £/min for this project),
the pressure drop across these tubes must be equal to or less
than a certain maximum value. This corresponds to the rated
capacity of the pump at a particular pressure drop. Sampling
tube parameters which affect the magnitude of the pressure drop
include:
Type of sorbent material packing
Mesh size of sorbent material packing
48
-------
Amount of glass wool plug
Diameter of the tube
Amount of sorbent material packing
Of the above parameters, the types and amounts of sorbent mate-
rial packings will be mainly determined by the capacity of these
sorbents for atmospheric pollutants under the anticipated sam-
pling conditions. The mesh size of the sorbent materials offers
some versatility, but there are limitations to the mesh sizes
available. The amount of glass wool used to contain the sorbent
material within the tube can be a critical factor in determining
the pressure drop, so that care should be taken to ensure that
the minimum amount of glass wool is used. In larger diameter
tubing, a method of restraining the glass wool plug may be nec-
essary when minimal amounts of glass wool are used. Tapering the
ends of the tube or inserting a stainless steel screen are two
possible ways to secure the glass wool plugs.
The experimental parameter that offers the most flexibility in
determining the pressure drop is the tube diameter. The only
restriction on the diameter of the tube is the necessity to
achieve rapid and uniform heating of the sorbent bed during the
thermal desorption process. This becomes increasingly difficult
with increasing sampling tube diameter.
The pressure drops across sampling tubes of various shapes, diam-
eters, and sorbent beddings were measured using a differential
pressure gauge. Flow rates were set by using a Brooks Model 5841
thermal mass flow sensor/controller. Owing to the low precision
of the differential pressure gauge used in this investigation,
the accuracies of the pressure drops measured below 5 psig are
questionable, and those measured below 1 psig are only very rough
approximations. An example of the type of data generated during
49
-------
these pressure studies is given in Table 16 and depicted in
Figure 19.
These pressure drop studies led to the conclusion that a 10-nun
diameter represents a reasonable compromise between the require-
ment to have the smallest possible diameter for efficient thermal
desorption and yet maintain an acceptable pressure drop across
sorbent-filled tubes. Two designs, straight and tapered, with
10-mm cross sections were evaluated, taking into account the
desirable design features of the miniature collection system.
Although the tapered tubes had higher pressure drops than the
straight tubes, the difference was not significant. Furthermore,
the tapered design offered advantages for the containment of the
sorbent since smaller amounts of glass wool can be used with no
"blow outs" of sorbent material. If one wished to reduce the
required amount of a glass wool even more, stainless steel
screens would be easier to place within the tapered tubes than
the straight tubes (which would probably have to be scored on
their inner surfaces). Finally, lighter, less expensive [0.6 cm
(1/4 in.)] fittings can be used with the tapered tubes, which is
an important factor in the design and construction of a miniature
collector. The specifications for the cartridge design determined
to be most desirable are given in Figure 20 along with a sketch of
a sampling tube.
~°-
6 cm
* i.o cm *
12 mmO.D. 10mm I.D.
~0.
r >
6 cm
6.5 mm
4mm 1.
0.
D.
i
D
GLASS WOOL SORBENT ;ATER,AL GLASS WOOL
Figure 20. Specifications of selected sampling tube design.
50
-------
rrj
CO
w
04
ID
E"1
O
12
H
-j
£H
£
CO
co
CO
f^
1 ^
o
K
o
f^
EH
rtj
Q
1
Q
3
ID
co
CO
§
fr
VD
rH
W
rt^
CQ
EJ
^
C
0
H
a
H
SH
U
tn
T3
d)
43
4->
4J
C
0
-~~ O
O T3 ^T
\ 3 C CM
O 1) w
cn .c! ,Q v£> cn
in 4-i M m td
f8 -H O \ rH
0
0
n ^-^
1 'O » go
CD ^ 3
43 -H kD rH
r4 tn cn w
0 rH ^D CO
tn 4J td
M O 0 \ rH
QJ C 3 Cn
§ ^
r^
»
S tn cn cn o
0) go
-X E co 3
td rH cn
a o o -w
id co CN co tn
O O O \ rH
CM in 3 Cn
*
U tn cn Cn o
O O 3 Cn
3 o
O 4-
rH n
PM (
43
Cn
-H
CO
a
-
<
> C
J -H
J g
3 \
co cn co o o
rH cn co ^f CM
CO rH
CM
rH
0 >£> rH
4J 1 1
0 0
rH
rH
O *£> ^Ji rH
n H o o o
e
co o co ^ co
si* cn co rH o
rH
n n rH CM O
O CN LO CN rH
CN rH
CM co ^ o m
co in I^D co rH
CM rH
in
in CN
rn CM rH O O
(
a
0
rfj Cn
H
o) cn
rH a
cn cn
tn
0)
a u>
Cn 3
H tn
cn co
a
cn
H C C
id -H 3
g ft go
0?
<* cn
rH ro cn
fc rrt i
fc ID *rH
U 4J W
Q co id a
id
o a >< CM
rH l£>
g CO £{
u o o
rH
VD CO CO O
X C CD M
0 O C 3
0^ *rH -i | CO
- 4-> rH CO
Cn
(0 X) O
51
-------
NOTE: lPSIG = 6.9x 103 Pa
30
20
10
//>'
,/ S S3
J ' .' ..-"
/*.'.' s&
/I-'."""
TENAX-PORAPAK
AMBERSORB
IN SERIES
/ / X
y 60180
/ I
/>TENAXGC
/ / 35/60
/ /
/ /
PORAPAK N
50/80
.-'
X
AMBERSORB XE-340
H
GLASS WOOL
-0/min
Figure 19.
Pressure drops vs. flow rates
for 15 cm long, 0.6 cm O.D.,
4 ram I.D. glass sampling tubes
52
-------
Evaluation of Sorbent Package
Five sorbent materials have been evaluated to determine their
basic properties, including special abilities and potential
problems.
Table 17 is a compilation of the major sorbent material proper-
ties evaluated. Along with the name of the adsorbent are included;
The manufacturer's stated temperature limit (or >400°C).
The temperature used for sorbent conditioning.
The temperature used for sorbent desorption.
The chemical composition of the polymer.
Some of the major thermal decomposition products.
A qualitative description of the background to be
expected from the sorbent material.
The pressure drop across tapered tubes containing
1.0 g of adsorbent at a 3 £/min flow rate.
A qualitative description of the type of compounds
that will potentially be quantitatively trapped by
the adsorbent (capacity >480 £/g).
A qualitative description of the type of compounds
that will potentially be quantitatively thermally
desorbed for analysis.
The range of utility based on 6 (described
previously in this report).
53
-------
r .
PH
^
a
u
w
H
EH
tf
Qj
0
«
CM
EH
W
«
cd
o
w
Q
, 1
Ed
«
EH
S.J
0 - ^
u °
1 ^ 1
IX 4-1 CL)
3 4J
C
o
H
a
0
o
a
>
-*
<0
a
3
fl XI
* cr
Q, -r-4
a
c
3
0 rH
^ a>
^ 0)
U r-l
CQ
H C
i 01 c > t/i c >> in
O r-| 13 OrH-O OrHT3 1
H QJ N C HO) C H QJ C (d
^ 0 T3 "§> £ ^ 0 "-0 "& E* ^ 0 TJ "& E 1
fd Q) rH Q) igQJrHQ) XI
QjrflCiflH QJiUCflH QJ £14-tl2"> Jq4J > > cx
1
>i 0) C >t Q) >i QJ
QJi-|iM OJ H (13 OJ-H^n)
rHtdfdrt rH(0(dojg r~t(d(OQ)o r-i
croo4JoiiM crao4-)nJ CTJO-Ptd c
HSJQJrH -H6S)-(0 HSJVjO H
HCDWU1 aW-HHOlOl 01 -HOJUlCO OJ -H
!!!!P! Hritu !!!iii! !
"DSSo-H^o "aSflo'rHitlo 'aMflOrHtio 's
s: x .c x:
LO en w LQ
o
O rH rH O
rH ^^
C
0) 3
> 0 Q)
Q M >
toil i en cj i CJ ja tj» o i i
-OOO l4JUOO>UOO(dO T3
C^ Ct^-iHfl ,4JrH^. C
2DQJC CJHLJCU rHOC tnd) SOC Z
QQJ>,2 Q)Hm2S83t>,SS'ui>, PS>iS Q
Q-uoicp 4J4-)cNt7iraA<^-'Wt^CM--'tn owity o
o S a* 8
OJ (/I H -H TJ
C rH r-l 0) r-l QJ Q)
OJO OOJC 011C >
N C HCQJ M C 0) 1-1
CO! UO'H S^O'H QJ
XlQJQja pj-HH aHH ^
CC QJr-tiH. QJrHr-l 0
r-tQJQJrH rHCOH rHCOH
>i IM N >i >i O l-i M >,OrHM 01
J^>iCr^ C'di-iln CTTS^U C
-H4J,>, H >, >, O
ft!ai >&,&< 3
1 QJ OJ QJ
-H C C C
>i QJ O O
QJ >. -H H 01
£ C rH rH N
aojoi r-io -HO H
H£T3 >-I>H >i-4 C
Qan CM CM o
1 1 X H >, -H >, XI
\oixo > a >cu w
CN S SO
o o o o o o
o in 4J CN in o
m rH (N r-l f)
O in in o
CN ro r*~ cv
(^ (N iH m
o o o o
o m en o
T M r-l ^
A
Ul Wl 01
U E E Z E "S
U ^ J4 O
i o (do fl o w
x VB a co a co M
C in ^ o ^ O XJ
OJ **> 0 in o in E
E-< a- ^ a. <
0
o
o
in
^
QJ
rH iH
r* i-( rH 01
QJ O 0 E
4-1 C
4-> r4 & Ll J3
7T "t:, ,1 *7i
jj tji td nj
>. a -H o c
r* O ^ H
XI TJ XI JZ QJ
3
Ot
rH
rH I t-|
nj g QJ ^
C ft) 0 -*
' ^H CP C
H r* QJ
£ rH -H U] -H
H S -P1? IM
£ en -H 3 QJ
2 QJ 0 fid 0)
O U T3
XI «J
R3 XI 0)
~ C
u u: 0
4J 01 O1 Q
O
U
C XI C 0>
H + rji o>
4J -iH o > 2 O
o
QJ rH 01 QJ
C >i C N
Q) C Q) H
r-t -H N C
>. > C Q
C
0 0 -( X5
P D1 11
-P i-i X 0
HO W
XI TJ XI .C
r-l i-(
r-l I 0
fl E «*-!
£ >^
C U -P t/l
' r-l £ C C
H H 3
>i 4J £ >4-l 0
X r-t -H fQ £
-51 > _< 8
2 S o S 'o
QJ 1
> M
Q U
XI ns
111
O en 0
P 01 C7>
u
H
(d
rH
nj
o
u J
> u
0. r-( M
54
-------
Of these five sorbents, a combination of three has been selected
for use in the miniature collection system. These three sorbents
are listed below along with the major reasons for their selection.
Tenax-GC - The only high temperature (400°C) adsorbent available
which allows the quantitative thermal desorption of
low-volatility organic compounds.
Porapak R - One of the highest capacity polymeric adsorbents with
a reasonable background level (better than Porapak N)
and with an overlap in range of utility (<$_.) with
Tenax-GC.
Ambersorb XE-340 - Less difficulty anticipated with the desorp-
tion of compounds of intermediate volatility,
fewer detrimental effects by water and reac-
tivity with collected samples than with char-
coal. Also, its range of utility (6 ) leaves
the smallest gap between polymeric and carbo-
naceous adsorbents in the types of compounds
collected.
These sorbent materials will be packed into three separate
tapered glass tubes for sampling, with flow being directed either
in series or in parallel. The types of compounds that are ex-
pected to be trapped and desorbed by this sorbent package are
described in Table 18 for the two modes possible. If sampling is
done with the sorbent tubes in series the arrangement will be
Tenax-GC at the air intake, Ambersorb XE-340 at the air exhaust
to the pump, and Porapak R in the middle. In terms of total
range of utility (5 ) of the sorbent package, it should quanti-
tatively trap and desorb compounds ranging from M50 to >2,800
(6 ) with a possible "qualitative only" gap at ^750 (6 ).
55
-------
CO
w
Q
O
5"1
O
3»
t"^
HH
o
W
O
^
U
P-i
EH
b^"l
w
CQ
OS
o
CO
PH
O
(_
O
1-1
p_l
^J
j
"I
H
CO
rH
H
CQ
EH
CD
o
S
r- I
0)
i-H
.-t
fO
(0
(fl
0)
H
OJ
C/1
1-J
o
w
Q
a
r-
XI
o
CO
0)
Q
ft
(C
£-
JJ
fi
a
£
c
CO
01
rH ^
XI 3 rH
1-t O O"1 rH
O ft C (fl
01 £ -H H
QJ O T3 4->
TJ O 3 in
>t i-H O ft
t rH C
0) 03 -r-t >i
> rH
H . - C *
P -P T3 O 'O
(TJ W QJ 0)
P O ft O ft
-P i-H nj O fC
C (t3 J-i -C J-t
(Q 4_> 4-) 4-»
a
rH
rH
IB i i a 01
SH 01 SH (fl Tl
a 01 rH (0 in C
(0 4-1 -H a 4-1 3
SH C 4-1 O
-P -H 10 T! ~ a
rH C dC E
>i 0) O (0 O O
i-H S > O O
0) O * rH
> 01 >* 01 v SH
H rH Tl 0)
4-1 TJ o) c s:
(8 G 4-) 3 SN 4-1
4-1 (8 (B O rH O
-H -H a rH
4-1 3 Tl 6 id rH
C O 01 O -H rH
(0 rH E U 4J (B
a
01
Tl
C >.
XI 3 rH
SH O DI -H
o a c IB
01 E -rl -H
01 O Tl 4-1
T! U 3 V4
rH (8
>. rH u a
rH H G
01 IB -H >,
> rH
H ~ - C
4-1 4J TJ O TJ
(fl 01 0) 01
p o a o> a
H g a 01 a
4J rH (fl O (fl
G (8 SH X SH
IB '4-14J4-1
a
rH
(Oi a 01
SH 01 I 10 TJ
a 01 rH SH SH G
SH G -P a O
>i 01 O C O O
rH g > IB O O
01 O rH
> yj >< ui v SH
4-1 T! 0) C -C
(fl C -P 3 >i -P
4-1 (fl Ifl O rH 0
H -H a rH
G 0 01 O -n rH
IB rH E O -P (fl
a
o
o
CO
CM
A
0
4J
O
CJ in
O cn
1 ,
TJ 0 rH
XI O) O rH
SH E T! 18
O rl C TJ -H
Ul 0) IB 0) -P
a c 01 a (0
H rH (fl a
N -H SH
rH 01 -P -P TJ
0) J3 (d C
> 4J rH >, IB
-rl O rH
4-1 14H > rH -
(fl O 10 01
H 4-1 rH 4J C
4-1 01 01 SH 3
C O -P (0 O
(8 g (0 a a
a
>,
rH rH r-
rH 01 tfP
(fl -P - O
Ifl Ul O
a -H T3 rH
(fl Tl C V
SH 01 3 ~
4-> E 0 SH
SH a >i 01
>l 01 6 rH ,G
rH 4J O rH 4-1
01 C O 10 O
> -H -rl
-H 0) 4-1 rH
P Tt rH SH rH
10 C -H Ifl Ifl
-P 10 4J a
H (o a
4-1 S rH Tl (0
C O O C SH
IB rH > (0 4-1
3
OI
rH II*
rH E C SH Tl
(fl O -H (0 0)
SH a a
SH ,C >. (0
O TJ En rH SH
Ul 01 3 c -P
01 a o o
TJ a SH ^
(8 X 01 dr
>, SH 4-1 01 O
rH 4-1 .X 0 O
01 10 £1 rH
> U) 01 -P V
H TJ SH ~--
P G XI G1
IB 3 G >.
P O X -rl rH
H a IB -a rH
P g G 3 18
C O 0) rH -H
Ifl U EH O -P
a
rH
rH 1 >.
IB (fl 1 rH
rH 01 rH
a o (B (fl
(0 > a 01 -r|
H .C 4J
4-1 >i Ul 4-1 SH
rH TJ IB
>, 0) C -C Q.
rH 4J 3 CH
01 IB Q 3 Tl
> -ri a o c
H TJ g SH ffl
P 0) O -G
IB g O -P -
4J SH X
-H 41 4) Oi 18
C C -H -H 01
(fl -rl 4J 01 EH
a
0
o
in
h
rH
f
O
p
On O
in
* ?
a
(fl -
SH E
O 'O
Q<
3
O
rH 01
01 03 XI
E T! C H
O C 3 0
Ul Ifl O 01
a oi
XI >i E Tl
SH *H 0
O 01 O >i
03 4-1 rH
0) Ifl >i 01
C -H 4-> >
T) -rl -H
~~ 0) rH -P
dP E -H Ifl
O SH -P 4-1
O 01 (0 -rl
rH 4-J rH -P
V C O C
' -H > (0
a
rH
rH
18
a
18
SH
-P
^,
rH
01
Ul >
Tl -rl
C -P
3 18
a H
S "-P
0 C
O 18
a
en
g
XI
SH
O
01
01
rH
01
-H
-p
(0
4-1
H
4-1
C
Ifl
a
01
, a (fl a
rH C 10
rH 01 01 rl
IfllT) EH 4-1
a 3 f. >
(fl 0 O1 rH
SH S, 3 01
P S O >
0 SH -rl
0 Si 4->
* 4J (0
O SH -P
O O O1 -H
V 4J -rl C
~^ O Ul (13
c
O
fl
1
[t
X
£3
S-
C
Ul
SH
Oi
£
<
^,
rH
r|
tP
-rl
0"
_f^
4-1
UH
O
-P
Ul
|
1
SH
01
-p
C
-H
TJ
C
(fl
Q
rH
0)
.C
4-1
UH
O
O
01
a
a
!c
CT
-H
JS
Oi
si
-p
UH
O
4J
o
c
£
0
4-1
o
in
rr
,?
rH XI
01 TJ 1 SH
> 41 >l E O
rl a rH 01 TJ 0 01
4-1 arH E C 0 01
Ifl IB (fl O IB Tl
4J SH -H U3 ^- 01 SH
H -P -P rH 4-1 01
4-1 SH -~ >i -n O .C
IB -H a O 01 10 O
3 rH O -P rH TJ
t^ifl *rHtOOG-P
H 01 V -H > (fl 01 01
01 -P TJ ~* T! O TJ
rHSHC OI^-EC
i-t (fl 3 X) E ^4 U3 3
(o aooiTJCd^a
OCOOlCOOi? O
> IB UTS-HX1 a^^O
>
rH 0)
UH rH SH
O 10 O
01 rl S
0) -P rH 4-1
rH Ul -H SH G
H 0 4-1 tfl 01
4-1 E 10 a > 01
10 rH 01 rH
rH TJ O TJ -H
o c > c a -P
$> (8 (B IB IB
>1 H rH .
>, Ul rH U) 4-1 O 01
rH Tl S. Tl > Tl
ID C BI C «> C
4-1 3 -H 3 * >i 3
IB O X 0 O rH 0
na a o x a
o g oi e H 01 g
01 O A 0 V -H o
E O 4-> U XI U
4J rH
11 G -rl Ifl
01 1 O rH -H
O E -n 4J
.C O SH 4-1 SH
JJ O 01 10 10 XI
Tl rH a SH
>, 01 SH O O
rH rH O > >1 U)
XI -H X) rH 01
rH 4-1 01 C TJ
01 ID 4-1 -P O
UJ rH (fl CO ~-.
O O J3 -H >i dr
a > -P 13 18 O
01 g 0
P >1 Ul E rH
a rH Tl H ff, V
0) X C 01 O
u cr, 3 4-1 -H
X rl O C C, >!
01 X a -H S rH
-P rH
SIB .
ij 1 -H TJ
TJ 01 111 g -P 01
C C S. K 0 SH a
a 3 o a IB
O TJ 01 SH
§SH C £ rH (P 4-1
XI (fl -P -H C
U 4-1 -H -p -H
X S (B <1> *
01 &i (8 rH A O
rH C C M O O
H -H 01 (8 > 01 i-H
4-1 x E-" a TJ v
(fl IB 10 01 C
O SH J3 O O O >i
> xi -P DJ g an
H
56
-------
TASK 3 - DESIGN OF THE PORTABLE MINIATURE COLLECTION SYSTEM
PROTOTYPE
There were a number of sampling configurations considered for use
in a miniature sampling system based upon the selected sorbent
materials (Tenax-GC, Porapak R, and Ambersorb XE-340). Consider-
ation has been given to both series and parallel arrangements of
the sampling tubes, and tubes of 12 mm O.D. (see Figure 20) and
17 mm O.D. were examined for this application. To determine the
total pressure drop requirements for these sampling configura-
tions, the information given in Table 19 (see also Section II.B.3)
was used in the following equations.
Series
Total AP = Tenax Tube AP + For. R Tube AP + Arab. 340 AP (5)
Parallel
Total AP = [(Tenax Tube AP)~1 + (Por. R Tube AP)~1
+ (Arab. 340 AP)-1]-1 (6)
The results of these calculations for different sampling config-
urations are given in Table 20. Note that the flow rate through
each sampling tube in the parallel arrangement is only one-third
of the total flow rate required through the pump. Therefore, the
major pump requirement for any sampling configuration is its
ability to pump at a certain flow rate over a particular pressure
drop. The information given in Table 19 is plotted in Figures 21
and 22 superimposed upon the pressure drop versus flow rate
(manufacturer's) specifications of a number of pumps. The pumps
which are described in Figure 21 are small, light-weight models
that are currently available on the market for miniature sampling
applications. Data on two larger, heavier pumps are given in
57
-------
o
!S
EH CO
W
CO PQ
CO D
O EH
PS
U O
TJ
CU Q
JH
3 O
CO
rn ^
UJ C
cu g
M
PU CN
H
0
4
n
s.
c
1
r.
H
fd
O
N
w
c,
H
«.
cu
^
rd
Cn
H
CO
p^
V
PU
o
td
O
fd
g
4-)
C
a)
Q
SH
O
CO
in
O ^I* CO *^f CO O^ ^f O4 C^ CO CP\ C^ V^ r-H C5
t>»
OrHCM iHOvl^D rH^T^ CNU3CO inHin
c? V A A rH rH CM
A
inin inm ininin in
O O i 1 OrHCN OrHCN i-HCNin CN'S'CTv
* *
ooo ooo ooo ooo ooo
V V A A A
n
C7>
C7> CT1 O
00 4->
^r . o o
ro H CN
1 rH M cu en ^
(U O X! X! > CU XI
43> SCO CO O-H CO
SnOJ CU UCU 43SH U
a
H t>i
-H
rH CU
4-1
rd" >i g
P-l r 1 *H
CO X
(0 3 O
0 O H
rH -H d,
> a,
X CD rd
SH
cri DJ to
H
0) PS
II G
H J^
in g fd
H SH DJ
en cu ro
Cu -P SH
0) O
rH P CU
rd 43 U
58
-------
CO
£?
EH
D
U
g
O
CJ
0
"
H
CM
^J
fn
UJ
EH
g
s
W
r
Q
rn
Cfl
O
PH
u
rH|
_
PH
§
Q
W
PH
tD
cn
W
PH
CM
-.
H
EH
0
EH
"
S
TABLE
-P
Q)
g
(U
tn
rd
M
rH
rd
(ji
C
H
rH
g
cn
rH
0)
rH
rH
(0
?
-P
nP
O
o) XI
K -
1
H H| N
C +
H
< rH | ^
H +
rd
P rH | X
Q | i
-P
II
CM
0)
X!
g P
g -P
H Q
0
cn
(1)
X)
i d
CN
rH Q
.
O
eries
cn
_rj
CJ1 v SH
^ P Q) -H
0 0 XI g
rH rH P \
Cn X! -P <=*
4J
rd
O
ffi
N
H +
fi >i
H
"^.^ _l
rH X
-P II
O
-P
CM
cn
(U
X)
13
rH Q
O
cn
D
X!
g P
g -P
CM
rH Q
O
0)
-p .<~i
CO tP » C
rH P QH-H
0 g g
O ,£ PH°?
rH -P
CM
r- m t~-
H ro vo
O O O rH
in vo in
n r> m
... 1
0 0 rH
cri in oo oo
-------
NOTE: UN lin.
=2.5xl02Pa;
lib- 0.45 kg
CAPITAL INVESTMENT
(PRICE INCLUDES PUMP,
BATTERY AND CHARGER)
©Du PONTP4000-A $635
0 BENDIX BOX 55HD $385
© BENDIX C115 $385
© RAC 2392-A $250
WEIGHT <5lb
12 mm O.D. TUBES
IN SERIES
\ 17 mm O.D. TUBES
IN SERIES
12 mm O.D. TUBES
IN PARALLEL
17mmO.D.JUBES
IN PARALLEL
0.5 1 2 3 4 5
FLOW RATE, 4/min
Figure 21. Plots of requirements for various sampling
arrangements and the manufacturer's specifi-
cations for four miniature pumps.
60
-------
Q_
O
DC
O
UJ
(ft
C/l
LU
300
250
200
150
100
50
10
CAPITAL INVESTMENT
CASE
4-6 hr BATTERY
CHARGER
~$175
-$100
~$ 35
PUMP (EITHER MODEL) ~$ 75
TOTAL ~$385
WEIGHT ~32 Ib
NOTE: I'm. H20=2.5x 1C2 Pa
1 lb.= 0.45kg
12mmO.D. TUBES
N SERIES
17mmO.D. TUBES IN SERIES
12 mmO.D. TUBES IN PARALLEL:
17mmO.D. TUBES IN PARALLEL
468
FLOW RATE, i/min
Figure 22. Plots of requirements for various sampling
arrangements and the manufacturer's specifi-
cations for two large (desk-top) pumps.
61
-------
Figure 22; these could be used in a "desk-top" sampling system.
Some capital investment and weight information is also supplied
in these figures.
Figure 23 is a rough sketch of a possible miniature sampling
system (one that would permit series or parallel sampling) and a
description of its most outstanding features. This system would
be small and light enough to attach to a person's belt, and would
operate for 8 hours. However, it would have its flows set by
delicate, easily broken needle valves and fairly inaccurate flow
meters, and would be more expensive than a larger sampling system
(see Figure 24). The maximum capabilities of this system are
also described in Figure 23 for the four sampling configurations
given in Figure 21. For example, with 12 mm O.D. sampling tubes
in parallel, a maximum of about 1 £/min could be passed through
each tube (3 5,/min through the pump) across the pressure drops
existing for that sampling configuration at that flow rate.
A larger, "desk-top" sampling system is similarly sketched and
described in Figure 24. This model would be too bulky and cum-
bersome to carry for an extended period, so it would be placed on
a stand, table, desk, or laboratory bench during sampling. The
design of this sampler offers a number of options. For example,
the power requirements may be supplied by one (lasts 4 to 6 hours)
or two (lasts >8 hours) batteries enclosed in the same case as
the pump and sampling tube holder or in a separate case. Power
requirements may alternatively be supplied by a transformer
adapted cord to a 120-V wall socket. Flow rates may be set by
needle valves and flow meters (as in the smaller sampling system)
or by critical orifices (which are more accurate and less easily
damaged, but require more pressure drop). The maximum capabili-
ties for the desk-top model are also described in Figure 24 for
the sampling configurations in Figure 22. Higher flow rates are
attainable for this model through the sorbent tubes which would
permit more sample to be collected. Whether or not this increase
62
-------
SAMPLING PORTS
(HOSES OPT.)
SAMPLING TUBE.
(3) HOLDER
FLOW METERS
PUMP, BATTERY
& CHARGER
NEEDLE VALVES
Weight
Cost
Size
Operational time
Method of setting flow
Durability problems
Capabilities (max)
(*preferred)
2.3 kg (<5 Ib)
$635 for DuPont P4000-A (pump, battery and charger only)
2 or 3 times the size of most pocket calculators
>8 hours
Needle valves and small flow meters
Needle valves fragile and pump near max for long
periods
tubes - 0.5 a/min
- Series, 12 mm O.D
- Series, 17 mm O.D. tubes - 2
*Parallel, 12 mm O.D. tubes -
Parallel, 17 mm O.D. tubes -
to 2.5 Jl/min
1 £/min
1 £/min
Figure 23. Sketch and description of small personnel sampler.
63
-------
OUTER CASE
SAMPLING TUBE (3)
HOLDER
t"-?~\ t~-/
BATTERY
(MAY HAVE 2)
fir
$±&
frS^&S
SAMPLING PORTS
(HOSES OPT.)
PUMP
Weight
Cost
Size
Operational time
Method of setting flow
Durability problems
Capabilities (max)
(*preferred)
M4.5 kg [>32 Ib; %27 kg (^60 Ib) if 2 batteries]
^$385 (for case, 1 battery, charger and pump only)
33 cm (13 in.) long, 20 cm (8 in.) wide, ^13 cm
(5 in.) deep
4-6 hr (>8 hr with 2 batteries or wall adapter)
- Needle valves and flow meters or critical orifices
- Needle valves (if used) fragile
- Series, 12 mm O.D. tubes - 4 to 5 £/min
Series, 17 mm O.D. tubes - 7 to 8 &/min
*Parallel, 12 mm O.D. tubes - 2.5 to 3 £/min
Parallel, 17 mm O.D. tubes - 3.5 to 4 £/min
Battery may be put into a separate pack and/or a 120-V line with transformer
may be used to adapt this sampler to a wall socket.
Due to pressure drop required at orifices the capabilities would be reduced.
Figure 24. Sketch and description of large desk-top sampler.
64
-------
in sample will significantly improve analytical sensitivity will
be investigated to determine if the desk-top model, with its dis-
advantages of cumbersome size and weight, should be used in this
project. The final design of the portable miniature collection
system will be completed during the second year of this contract
as part of Phase II, in order to permit discussion of design op-
tions with the EPA.
ADDITIONAL WORK - ANALYTICAL DEVELOPMENT
During this contract year a Hewlett-Packard Model 5840A GC with
glass capillary option was purchased for use in the development
of analytical techniques during Phase II of this research pro-
gram. After the initial setup of this instrument, a Perkin-
Elmer, 100-meter SF96 WCOT column was installed. To demonstrate
the analytical capabilities of this system, samples of a highly
complex organic mixture (JP-4 jet fuel) were analyzed. Figure 25
contains two chromatograms. Chromatogram A is from an analysis
of JP-4 done previously with a packed column. Chromatogram B is
an analysis of JP-4 done on the capillary GC system at a flow
rate of 1.2 m£/min of nitrogen and a sample split of about 200 to
1. The increased resolution of the capillary column is readily
apparent.
Work was continued in the area of analytical development with the
design, construction, and installation of a special capillary
inlet system. This system was made to permit the thermal de-
sorption of sampling tubes, reconcentration of the desorbed
sample in a trap, and introduction of the sample into the capil-
lary GC for analysis. In order to install this special capillary
inlet system (desorption chamber, valve, and trap) it was first
necessary to make two modifications to the HP 5840 GC. The first
modification involved the installation of an auxiliary "desorp-
tion" carrier for use during the thermal desorption of sorbent
tubes. This auxiliary desorption carrier was directed through an
65
-------
A. JP-4 separated by a packed column.
-JW
i
i.i
1
J^M^
14
B. JP-4 separated by a SF96 WCOT column.
Figure 25.
Two chromatograms of JP-4 jet fuel. Numbered peaks
are the n-alkanes of corresponding carbon number.
66
-------
available flow sensor on the 5840 so that the flow could be
monitored directly. This modification is shown in Figure 26, A.
The second modification (Figure 26, B) involved the installation
of a three-way [3.2 mm (1/8 in.) swage] ball valve within the
capillary flow system to allow for the selection of either a
"normal injection" flow pattern or a "purge trap" flow pattern.
In the "purge trap" flow pattern a source of carrier is diverted
from the normal flow pattern and directed into the capillary
inlet system. This provides a source of carrier for sweeping the
capillary trap when it is switched into the analyze position
(i.e., into the flow system with the capillary column).
The current configuration of the capillary inlet system is
sketched in Figure 27. This inlet system performs three major
functions. The first function is the thermal desorption of sor-
bent tubes and is accomplished by a copper tube furnace wrapped
with Nichrome heating wire. The temperature of this tube furnace
is controlled by a Variac and read with a digital voltmeter.
Flow selection and direction is the second major function of the
inlet system. This is accomplished with a six-port, two-position
valve within a heated oven compartment. The other function of
cryogenic trapping and focusing of the desorbed sample is accom-
plished in a 1-m, 0.8 mm (0.03-in.) I.D. transaxially coiled
nickel trap that is cooled by liquid nitrogen during sample
collection and heated by electrical resistance during sample
introduction to the capillary column. Thermocouples are attached
to a pyrometer to permit monitoring the temperatures of the
valve, valve oven, and trap.
A flow diagram of the capillary inlet system is shown in Figure 28
for the two valve positions. In the "trap" position a sorbent
tube is desorbed by the tube furnace and its sample is swept by
the auxiliary "desorption" carrier through the valve into the
nickel capillary trap, which is 50% submerged in liquid nitrogen.
The flow out of the trap is passed through the "A" injection port
67
-------
r
FLOW
CONTROLLER
FLOW
SENSOR
AUXILIARY
"DESORPTION" CARRIER
-A CARRIER
AUXILIARY
"DESORPTION" CARRIER
TO CAPILLARY
"INLET SYSTEM
A. Auxiliary "desorption" carrier modification.
SPLIT MODE
TOGGLE ^ TOGGLE
VALVE 1 VALVE 2
AUXILIARY
PURGE TRAP"
CARRIER
NV2
INJECTION PORT
TO CAPILLARY
INLET SYSTEM
FROM FLOW CONTROLLER
FLOW MODULE
3-WAY BALL VALVE
SELECTS TRAP FLOW
OR NORMAL OPERATION
B. Auxiliary "purge trap" modification.
Figure 26. Flow schematics of two GC modifications
68
-------
o
O
oo
o
<
UJ
Q_
<
O
°z TT
LU t-;
LU O _l
£ =J^<
O < cvj >
> u_-!i!
<
X
CO
z
o:
i
>-
_i
Q_
O
CO
LU
Q
Q OC
LU O
£^
LU v£>
=> LU Q.
-J ti <
O 5: Qi
> ^ l-
§
-p
w
>l
en
-P
Q)
H
(0
H
H
H
a,
o
x:
4->
-H
s
H
&
tO
U
r-
CN
0)
(31
H
©
©
69
-------
AUXILIARY "DESORPTION"
CARRIER (-30 mi/mini
TRAP MODE
SAMPLE DESORBED FROM
SORBENTTUBE INTO TRAP
SORBENT
TUBE
=ff-HEAT « AUXILIARY "PURGE
TRAP" CAPILLARY
CARRIER (~lm£/min)
CAPILLARY "A"
INJECTION PORT INJECTION PORT
AUXILIARY "DESORPTION"
CARRIER (~30m.e/min)
ANALYZE MODE
SAMPLE DESORBED FROM TRAP
INTO CAPILLARY COLUMN
AUXILIARY " PURGE
TRAP" CAPILLARY
CARRIER (~1 m-E/min)
CAPILLARY "A"
INJECTION PORT INJECTION PORT
Figure 28. Flow schematic of capillary inlet system.
70
-------
to the "A" FID for monitoring trap breakthrough. When the valve
is switched to the "analyze" position, simultaneously the tube
furnace is cooled, the liquid nitrogen is removed from the trap,
and the trap is heated. The auxiliary "purge trap" flow then
sweeps the collected sample out of the trap and into the capil-
lary injection port. When the capillary GC is operated in the
"split" mode, most of the sample entering the injection port is
vented, and only a small portion enters the capillary column for
chromatographic separation and analysis. A modification has been
made to re-collect the vented portion of a sample onto another
sorbent tube for additional, subsequent analyses. A more refined,
neatly packaged, and conveniently operated version of the capil-
lary inlet system is anticipated after evaluations and modifica-
tions are completed.
Some preliminary evaluations of the capillary inlet system on the
capillary GC were conducted to determine the utility of this
system. To permit initial evaluations of this system by sample
injection, a 0.6 cm (1/4 in.) swage tee was installed at the base
of the tube furnace, capped with a septum nut, and wrapped with
heating tape. For all capillary inlet system evaluations, the
flow out of the trap was monitored by the "A" FID during sample
trapping to assure that no breakthrough occurred. The "B" FID
(glass capillary) was monitored during sample analysis. For
injected samples of ether, no breakthrough was noted on the "A"
FID when the nickel capillary trap was cooled, but immediate
breakthrough occured when it was not cooled. Similar samples of
pentane and hexane were also analyzed with optimistic results.
Injected samples of JP-4 were then analyzed for comparison with
earlier, regularly injected samples (see Figure 25). Two nickel
traps were compared in these analyses which were of different,
[0.8 and 1.6 mm (0.03 and 0.04 in.)] internal diameter but other-
wise identical. Both analyses indicated slight peak broadening
and poorer resolution compared to previous samples, but of the
two traps, the 0.8 mm (0.03 in.) I.D. trap was judged to yield
71
-------
slightly better results. A chromatogram of a JP-4 analysis with
the smaller diameter trap is shown in Figure 29.
Next, a six-component mixture containing pentane, hexane, heptane,
octane, benzene, and toluene was obtained and evaluations con-
ducted with this mixture. The analysis of an injected sample of
this mixture is depicted in chromatogram A of Figure 30. Addi-
tionally, samples of this six-component mixture were generated
using Monsanto Research Corporation's sample generation system
(see Appendix A) and collected on Ambersorb XE-340 sorbent sam-
pling tubes (1 g of sorbent). An analysis of a l-£/min, 16-hr
sample is shown in Figure 30, B, and was accomplished using a 30-
min sorbent tube desorption with nickel capillary trap collection
and a 5-min nickel capillary trap desorption. For some sorbent
tube samples, the vented (500:1 split) portions were re-collected
on other sorbent tubes at the capillary GC splitter. The analysis
of one of these re-collected samples is shown in Figure 30, C.
(The additional peaks in Figure 30, C, are from residual contam-
ination at the splitter vent outlet.) Re-collected samples
showed an average of 93% of the original sample amount if analyzed
the same day as collection. Re-collected samples stored for a
day or two showed between 40% and 67% of the original sample,
with refrigeration showing no effect.
1
Figure 29.
J
10 11
15
U*
-------
I
-
4 "*
- 2 ?
A. Injected sample of six-component mixture.
U
* SI
B. Sample of six-component mixture from an Ambersorb XE-340
sorbent tube generated by the sample generation system.
Figure 30.
Evaluation of capillary inlet system with
samples of a six-component mixture (pentane,
hexane, benzene, heptane, toluene, and oc-
tane) , collection in liquid N2 cooled trap, and
desorption into a SF96 WCOT column for analysis.
73
-------
C. Sample of six-component mixture from an Ambersorb XE-340
sorbent tube collected from the capillary GC splitter.
Figure 30 (continued)
74
-------
REFERENCES
1. Holzer, G., H. Shanfield, A. Zlatkis, W. Bertsch, P. Juarez,
H. Mayfield, and H. M. Lieblich. Collection and Analysis
of Trace Organic Emissions from Natural Sources. Journal of
Chromatography, 142:755-764, 1977.
2. Pellizzari, E. D. , J. E*. Bunch, R. E. Berkley, and J. McRae.
Collection and Analysis of Trace Organic Vapor Pollutants in
Ambient Atmospheres. The Performance of a Tenax GC Car-
tridge Sampler for Hazardous Vapors. Analytical Letters,
9(1):45-63, 1976.
3. Pellizzari, E. D., R. H. Carpenter, J. E. Bunch, and
E. Sawicki. Collection and Analysis of Trace Organic Vapor
Pollutants in Ambient Atmospheres. Thermal Desorption of
Organic Vapors from Sorbent Media. Environmental Science &
Technology, 9 (6):556-560, 1975.
4. Gallant, R. F., J. W. King, P. L. Lewis, and J. F. Piecewicz.
Characterization of Sorbent Resins for Use in Environmental
Sampling. EPA-600/7-78-054, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1978.
151 pp.
5. Burrell, H., and B. Immergut. Solubility Parameter Values.
In: Polymer Handbook, J. Brandup and E. H. Immergut, eds.
Interscience Publishers, New York, New York, 1966.
pp. IV-341 - IV-368.
6. Small, P. A. Some Factors Affecting the Solubility of
Polymers. Journal of Applied Chemistry, 3(2):71-70, 1953.
7. Adams, J., K. Menzies, and P. Levins. Selection and Evalu-
ation of Sorbent Resins for the Collection of Organic Com-
pounds. EPA-600/7-77-044, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1977.
61 pp.
75
-------
APPENDIX A
STANDARD SAMPLE GENERATION SYSTEM
In order to evaluate the sorbent sampling systems used in this
project it is necessary to have a method for generating known
concentrations of organic compounds in dynamic gas streams. MRC
has developed a dynamic standard sample generation system based
on the controlled syringe injection of organic liquids into a
flowing stream of gas (N2)/ followed by vaporization and subse-
quent dilution.
Figure A-l is a schematic diagram of the sample generation sys-
tem. The main frame of the system is a modified F&M Model 700
gas chromatograph. The chromatograph has been stripped of the
protective metal covering and the detector has been removed. The
oven [30.5 cm x 30.5 cm (12 in. x 12 in.)] was relocated from the
right to the left side of the main frame. Four heated zones were
available on the original GC. One of these zones controls the
oven temperature while the other three are available for heating
various components of the sample generation -system. Two of these
are used to separately control the temperature of the two three-
port injection blocks (see Figure A-l). The final zone is avail-
able for heating transfer lines for direct interfacing with a
detector for frontal analysis capacity studies. A Simpson pyro-
meter (0°C to 500°C) has been added along with a selection switch
and necessary circuit modifications to allow the monitoring of
temperatures in the four zones.
76
-------
-------
A 45.7 cm x 45.7 cm x 0.6 cm (18 in. x 18 in. x 1/4 in.) alumi-
num plate has been attached to the mainframe to the right of the
oven to accommodate the syringe drives. The necessary bulkhead
fittings, tees, toggle valves, needle valves, tubing, pressure
gauges, and rotameters were also added to accomplish the config-
uration indicated in Figure A-l. Materials of construction for
the flow system are either stainless steel or nickel. The entire
system is contained ina2.4mxl.2mxl.5m (8 ft x 4 ft x
5 ft) exhaust hood with access at both front and back through
double sliding doors. The flow control/sensing instrumentation,
pressure gauges, and rotameters are mounted on a panel and rack
assembly in the hood above the generation system.
The generation system functions in the following manner. A
source of carrier gas (N2) is introduced and split into primary,
secondary, and tertiary flows. The primary flow passes through a
Brooks Model 5841 mass flow sensor/controller which maintains the
flow at a preset value (0 to 1,000 mJl/min) . This flow passes
through individually heated and controlled injection blocks which
are designed to each accept three syringes mounted on Sage Model
355 variable control syringe drives. This allows for simultane-
ous introduction of six pure liquid components or potentially
many more if mixtures of compounds are used in the syringes. The
rate of injection can be varied from submicroliter-per-hour to
milliliter-per-minute rates by choice of syringe size and drive
rate. The liquid samples are vaporized in the injection blocks
and swept with the carrier gas into a constant temperature oven.
The major portion of the primary flow passes through the oven,
through a needle valve used to regulate the pressure in the pri-
mary flow path, and is expelled to a vent (VI). A small portion
(usually 'vlO m£/min) of the primary flow is split off through a
dual control needle valve and passes into a switching valve
(SVl). This valve is actually one-half of an eight-port switch-
ing valve. The second half is used for an identical function in
78
-------
the second dilution stage. The switching valve makes it possible
to switch a Brooks Model 5810 (0 to 100 m£/min) mass flow sensor
into the flow path (dashed line in Figure A-l) to measure the
exact split on the primary flow. Once the split has been estab-
lished, the flow sensor is switched out of the system so that the
sensing elements are not subjected to sample containing organic
vapors. The split is maintained by carefully controlling the
pressures in each of the dilution stages.
The pressures are monitored with Heise Model CM precision pres-
sure gauges [0 to 6.9 x 105 Pa (0 psi to 100 psi)]. By maintain-
ing a constant pressure in each of the dilution stages a constant
pressure differential is established between the stages assuring
a consistent split ratio across the needle valves.
The portion of the primary flow that is split out passes from
the switching valve and is combined with the secondary carrier
flow which enters the constant temperature oven through a Brooks
Model 5841 mass flow sensor/controller (0 to 2,000 m£/min). As
in the case of the primary flow, the major portion of the second-
ary flow passes through the oven and is expelled to vent V2
through a needle valve. A small portion (usually 10 mJl/min) of
the secondary flow is split off through a dual control needle
and passes into a switching valve (SV2). This valve functions
in an identical manner to valve SVl to place a second Brooks
Model 5810 (0 to 100 m£/min) mass flow sensor into the flow path
for measuring the second split.
The portion of the secondary flow that is split out passes from
the switching valve and is combined with the final dilution
stream which enters the constant temperature oven through a
Brooks Model 5841 mass flow sensor/controller (0 to 50£/min)
This final concentration is available for sampling through a
five-port manifold system that will accommodate sorbent sampling
tubes. The sampling rates through the sorbent tubes are maintained
79
-------
by Brooks Model 1110 rotameters at 'the exit of the tubes. The
unused portion of the final dilution stage is expelled to vent
V3.
With the sample generation system in this configuration it is
theoretically possible to achieve dynamic sample generation of
pure compounds down to low ppt concentrations. For example, a
concentration of M ppt should be achievable using pure benzene
in a 25-yJl syringe at a syringe drive rate of 25% of the 1/1,000
of full range setting and dilution stage flows of 1,000 m£/min,
2,000 m£/min, and 50 £/min, respectively, using 10 m£/min splits
from both the first and second dilution stages. Under these con-
ditions, the syringe drive mechanism delivers 0.0143 y£
80
-------
TECHNICAL REPORT DATA
(Plccsc read Iitiinn nons on the re, <.r*' b< tort i
RECORT NO
EPA-600/2-80-026
4 TITLE A\D SUBTITLE
PORTABLE MINIATURE SAMPLER FOR POTENTIAL AIRBORNE
CARCINOGENS IN MICROENVIRONMENTS
Phase 1. Development
6 PERFORMING ORGANIZATION CODE
3 RECIPIENT'S ACCESSiOf*NO
5 REJa°nRuaDryTE1980
i AUTHOR.S
8 PERFORMING ORG AN I Z AT I Or,
J. J. Brooks and D. S. West
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
Dayton, OH 45407
10 PROGRAM ELEMENT NO
1HE775 CB-004 (FY-79)
11 CONTRACT/GRANT NO
68-02-2774
12 SPONSORING AGENCY N AME AN D A DDRESS
Environmental Sciences Research Laboratory - RTP,NC
Office of Research and Development
U. S. Environmental Protection Agency
Research Triangle Park, NC 27711
13 TYPE OF REPORT AND PERIOD COVERED
Interim 9/77-10/78
14 SP'ONSORIN'G AGENCY CODE
EPA/600/09
15 SUPPLEMENTARY NOTES
16 ABSTRACT
A 3-year research project was initiated to develop a portable, miniature,
sorbent-type collection system for sampling and preconcentrating organics in
general, and carcinogens and associated compounds (e.g., mutagens, precarcinogens,
and cofactors) in particular, from ambient air. The purpose of such a system
is to assess the exposure of individuals and/or small groups of individuals
to these types of compounds in various environments. Inherent in the ability
to assess exposures is not only the sampling capability but also analytical
confirmation. The determinative step in this project will be capillary gas
chromatography/mass spectroscopy.
Progress during the first year was discussed and concerned the selection
of candidate sorbent materials; the selection of test compounds for sorbent
evaluation; the evaluation of the sorbent materials in terms of capacity,
desorption properties, and physical properties that relate to pressure drops
and ultimate system design, and the selection of a three-sorbent system based
on Tenax-GC, Porapak R, and Ambersorb.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS'OPEN ENDED TERMS
COSATI } icid Group
*Air pollution
*0rganic compounds
*Carcinogens
*Samplers
Portable equipment
Miniaturization
*Development
13B
07C
06E
14B
13M
18 DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
EPA Form 2220-1 (9-73)
19 SECURITY CLASS i Tins Report)
UNCLASSIFIED
21 NO OF PAGES
91
20 SECURITY CLASS -Tins
UNCLASSIFIED
22 PRICE
81
------- |